Movatterモバイル変換


[0]ホーム

URL:


はてラボはてな匿名ダイアリー
ようこそ ゲスト さんログインユーザー登録

「log」を含む日記RSS

はてなキーワード:logとは

次の25件>

2025-10-24

https://shorturl.asia/XUEfm

Log in to Charles Schwab ... Tradingbeyond the bell?Nowyou canaccess evenmore equities24/5. Tradeover 800 popular stocks and ETFs around the clockat Charles Schwab

Charles SchwabLogin Guide – Learn how to securelyaccessyour account, manage investments, and trade commission-free stocks and ETFs with ease.

Step-by-Step Charles SchwabLogin Process – A complete guide tologging intoyour Schwab account, checking real-time market updates, and managingyour portfolio.

‍Charles Schwab SecureLogin Guide – Follow oursimplelogin instructions toaccess investment insights, equity ratings, and financial planning tools.

How toLog into Charles Schwab – Abeginner-friendly guide to navigatingyour Schwab account, exploring fund options, and making informed investment decisions.

‍Charles SchwabLogin Help – Troubleshootloginissues,resetyourpassword, and gainaccess to commission-free trading and retirement planning features.

Login to Charles Schwab & Start Investing – Explore Schwab’s brokerage services, from mutual funds to ETFs, and secureyour financial futuretoday.

‍Charles SchwabLogin & Account Management –Sign in effortlessly to manageyour assets, track investments, andstay informed with market insights.

AccessYour Schwab Account – Complete the Charles Schwablogin process and unlock powerful investment tools, stock ratings, and retirement strategies.

‍Charles Schwab AccountLogin Explained – Asimple walkthrough for investors toaccess their accounts,review financialnews, and trade with confidence.

‍Charles SchwabLogin forBeginners – Start investingtoday with our easy-to-followlogin guide, plustipson portfolio management and fund selection

Disclaimer: Thiswebsiteis not affiliated with, endorsedby, or associated with Charles Schwab & Co., Inc.All trademarks,logos, and brand names mentioned are the property of their respective owners. The content providedhereis for informational purposesonly and should not be considered financial or investment advice. Users should visit theofficial Charles Schwabwebsite for accurate and up-to-date details regarding their services.

Permalink |記事への反応(0) | 17:18

このエントリーをはてなブックマークに追加ツイートシェア

https://shorturl.asia/HQdpL

Log in to Charles Schwab ... Tradingbeyond the bell?Nowyou canaccess evenmore equities24/5. Tradeover 800 popular stocks and ETFs around the clockat Charles Schwab

Charles SchwabLogin Guide – Learn how to securelyaccessyour account, manage investments, and trade commission-free stocks and ETFs with ease.

Step-by-Step Charles SchwabLogin Process – A complete guide tologging intoyour Schwab account, checking real-time market updates, and managingyour portfolio.

‍Charles Schwab SecureLogin Guide – Follow oursimplelogin instructions toaccess investment insights, equity ratings, and financial planning tools.

How toLog into Charles Schwab – Abeginner-friendly guide to navigatingyour Schwab account, exploring fund options, and making informed investment decisions.

‍Charles SchwabLogin Help – Troubleshootloginissues,resetyourpassword, and gainaccess to commission-free trading and retirement planning features.

Login to Charles Schwab & Start Investing – Explore Schwab’s brokerage services, from mutual funds to ETFs, and secureyour financial futuretoday.

‍Charles SchwabLogin & Account Management –Sign in effortlessly to manageyour assets, track investments, andstay informed with market insights.

AccessYour Schwab Account – Complete the Charles Schwablogin process and unlock powerful investment tools, stock ratings, and retirement strategies.

‍Charles Schwab AccountLogin Explained – Asimple walkthrough for investors toaccess their accounts,review financialnews, and trade with confidence.

‍Charles SchwabLogin forBeginners – Start investingtoday with our easy-to-followlogin guide, plustipson portfolio management and fund selection

Disclaimer: Thiswebsiteis not affiliated with, endorsedby, or associated with Charles Schwab & Co., Inc.All trademarks,logos, and brand names mentioned are the property of their respective owners. The content providedhereis for informational purposesonly and should not be considered financial or investment advice. Users should visit theofficial Charles Schwabwebsite for accurate and up-to-date details regarding their services.

Permalink |記事への反応(0) | 17:14

このエントリーをはてなブックマークに追加ツイートシェア

https://shorturl.asia/x8v1U

Log in to Charles Schwab ... Tradingbeyond the bell?Nowyou canaccess evenmore equities24/5. Tradeover 800 popular stocks and ETFs around the clockat Charles Schwab

Charles SchwabLogin Guide – Learn how to securelyaccessyour account, manage investments, and trade commission-free stocks and ETFs with ease.

Step-by-Step Charles SchwabLogin Process – A complete guide tologging intoyour Schwab account, checking real-time market updates, and managingyour portfolio.

‍Charles Schwab SecureLogin Guide – Follow oursimplelogin instructions toaccess investment insights, equity ratings, and financial planning tools.

How toLog into Charles Schwab – Abeginner-friendly guide to navigatingyour Schwab account, exploring fund options, and making informed investment decisions.

‍Charles SchwabLogin Help – Troubleshootlogin issues,resetyourpassword, and gainaccess to commission-free trading and retirement planning features.

Login to Charles Schwab & Start Investing – Explore Schwab’s brokerage services, from mutual funds to ETFs, and secureyour financial futuretoday.

‍Charles SchwabLogin & Account Management –Sign in effortlessly to manageyour assets, track investments, andstay informed with market insights.

AccessYour Schwab Account – Complete the Charles Schwablogin process and unlock powerful investment tools, stock ratings, and retirement strategies.

‍Charles Schwab AccountLogin Explained – Asimple walkthrough for investors toaccess their accounts,review financialnews, and trade with confidence.

‍Charles SchwabLogin forBeginners – Start investingtoday with our easy-to-followlogin guide, plustipson portfolio management and fund selection

Permalink |記事への反応(0) | 17:14

このエントリーをはてなブックマークに追加ツイートシェア

https://log-io-charles.webflow.io/

Log in to Charles Schwab ... Tradingbeyond the bell?Nowyou canaccess evenmore equities24/5. Tradeover 800 popular stocks and ETFs around the clockat Charles Schwab

Charles SchwabLogin Guide – Learn how to securelyaccessyour account, manage investments, and trade commission-free stocks and ETFs with ease.

Step-by-Step Charles SchwabLogin Process – A complete guide tologging intoyour Schwab account, checking real-time market updates, and managingyour portfolio.

‍Charles Schwab SecureLogin Guide – Follow oursimplelogin instructions toaccess investment insights, equity ratings, and financial planning tools.

How toLog into Charles Schwab – Abeginner-friendly guide to navigatingyour Schwab account, exploring fund options, and making informed investment decisions.

‍Charles SchwabLogin Help – Troubleshootloginissues,resetyourpassword, and gainaccess to commission-free trading and retirement planning features.

Login to Charles Schwab & Start Investing – Explore Schwab’s brokerage services, from mutual funds to ETFs, and secureyour financial futuretoday.

‍Charles SchwabLogin & Account Management –Sign in effortlessly to manageyour assets, track investments, andstay informed with market insights.

AccessYour Schwab Account – Complete the Charles Schwablogin process and unlock powerful investment tools, stock ratings, and retirement strategies.

‍Charles Schwab AccountLogin Explained – Asimple walkthrough for investors toaccess their accounts,review financialnews, and trade with confidence.

‍Charles SchwabLogin forBeginners – Start investingtoday with our easy-to-followlogin guide, plustipson portfolio management and fund selection

Disclaimer: Thiswebsiteis not affiliated with, endorsedby, or associated with Charles Schwab & Co., Inc.All trademarks,logos, and brand names mentioned are the property of their respective owners. The content providedhereis for informational purposesonly and should not be considered financial or investment advice. Users should visit theofficial Charles Schwabwebsite for accurate and up-to-date details regarding their services.

Permalink |記事への反応(0) | 17:07

このエントリーをはてなブックマークに追加ツイートシェア

https://shorturl.asia/F8dmk

Log in to Charles Schwab ... Tradingbeyond the bell?Nowyou canaccess evenmore equities24/5. Tradeover 800 popular stocks and ETFs around the clockat Charles Schwab

Charles SchwabLogin Guide – Learn how to securelyaccessyour account, manage investments, and trade commission-free stocks and ETFs with ease.

Step-by-Step Charles SchwabLogin Process – A complete guide tologging intoyour Schwab account, checking real-time market updates, and managingyour portfolio.

‍Charles Schwab SecureLogin Guide – Follow oursimplelogin instructions toaccess investment insights, equity ratings, and financial planning tools.

How toLog into Charles Schwab – Abeginner-friendly guide to navigatingyour Schwab account, exploring fund options, and making informed investment decisions.

‍Charles SchwabLogin Help – Troubleshootloginissues,resetyourpassword, and gainaccess to commission-free trading and retirement planning features.

Login to Charles Schwab & Start Investing – Explore Schwab’s brokerage services, from mutual funds to ETFs, and secureyour financial futuretoday.

‍Charles SchwabLogin & Account Management –Sign in effortlessly to manageyour assets, track investments, andstay informed with market insights.

AccessYour Schwab Account – Complete the Charles Schwablogin process and unlock powerful investment tools, stock ratings, and retirement strategies.

‍Charles Schwab AccountLogin Explained – Asimple walkthrough for investors toaccess their accounts,review financialnews, and trade with confidence.

‍Charles SchwabLogin forBeginners – Start investingtoday with our easy-to-followlogin guide, plustipson portfolio management and fund selection

Disclaimer: Thiswebsiteis not affiliated with, endorsedby, or associated with Charles Schwab & Co., Inc.All trademarks,logos, and brand names mentioned are the property of their respective owners. The content providedhereis for informational purposesonly and should not be considered financial or investment advice. Users should visit theofficial Charles Schwabwebsite for accurate and up-to-date details regarding their services.

https://shorturl.asia/F8dmk

https://shorturl.asia/F8dmk

https://shorturl.asia/F8dmk

https://shorturl.asia/F8dmk

https://shorturl.asia/F8dmk

https://shorturl.asia/F8dmk

Permalink |記事への反応(0) | 17:04

このエントリーをはてなブックマークに追加ツイートシェア

https://log-io-charles.webflow.io/

Log in to Charles Schwab ... Tradingbeyond the bell?Nowyou canaccess evenmore equities24/5. Tradeover 800 popular stocks and ETFs around the clockat Charles Schwab

Charles SchwabLogin Guide – Learn how to securelyaccessyour account, manage investments, and trade commission-free stocks and ETFs with ease.

Step-by-Step Charles SchwabLogin Process – A complete guide tologging intoyour Schwab account, checking real-time market updates, and managingyour portfolio.

‍Charles Schwab SecureLogin Guide – Follow oursimplelogin instructions toaccess investment insights, equity ratings, and financial planning tools.

How toLog into Charles Schwab – Abeginner-friendly guide to navigatingyour Schwab account, exploring fund options, and making informed investment decisions.

‍Charles SchwabLogin Help – Troubleshootloginissues,resetyourpassword, and gainaccess to commission-free trading and retirement planning features.

Login to Charles Schwab & Start Investing – Explore Schwab’s brokerage services, from mutual funds to ETFs, and secureyour financial futuretoday.

‍Charles SchwabLogin & Account Management –Sign in effortlessly to manageyour assets, track investments, andstay informed with market insights.

AccessYour Schwab Account – Complete the Charles Schwablogin process and unlock powerful investment tools, stock ratings, and retirement strategies.

‍Charles Schwab AccountLogin Explained – Asimple walkthrough for investors toaccess their accounts,review financialnews, and trade with confidence.

‍Charles SchwabLogin forBeginners – Start investingtoday with our easy-to-followlogin guide, plustipson portfolio management and fund selection

Disclaimer: Thiswebsiteis not affiliated with, endorsedby, or associated with Charles Schwab & Co., Inc.All trademarks,logos, and brand names mentioned are the property of their respective owners. The content providedhereis for informational purposesonly and should not be considered financial or investment advice. Users should visit theofficial Charles Schwabwebsite for accurate and up-to-date details regarding their services.

https://log-io-charles.webflow.io/

https://log-io-charles.webflow.io/

https://log-io-charles.webflow.io/

https://log-io-charles.webflow.io/

Permalink |記事への反応(0) | 17:02

このエントリーをはてなブックマークに追加ツイートシェア

https://log-io-charles.webflow.io/

Log in to Charles Schwab ... Tradingbeyond the bell?Nowyou canaccess evenmore equities24/5. Tradeover 800 popular stocks and ETFs around the clockat Charles Schwab

Charles SchwabLogin Guide – Learn how to securelyaccessyour account, manage investments, and trade commission-free stocks and ETFs with ease.

Step-by-Step Charles SchwabLogin Process – A complete guide tologging intoyour Schwab account, checking real-time market updates, and managingyour portfolio.

‍Charles Schwab SecureLogin Guide – Follow oursimplelogin instructions toaccess investment insights, equity ratings, and financial planning tools.

How toLog into Charles Schwab – Abeginner-friendly guide to navigatingyour Schwab account, exploring fund options, and making informed investment decisions.

‍Charles SchwabLogin Help – Troubleshootloginissues,resetyourpassword, and gainaccess to commission-free trading and retirement planning features.

Login to Charles Schwab & Start Investing – Explore Schwab’s brokerage services, from mutual funds to ETFs, and secureyour financial futuretoday.

‍Charles SchwabLogin & Account Management –Sign in effortlessly to manageyour assets, track investments, andstay informed with market insights.

AccessYour Schwab Account – Complete the Charles Schwablogin process and unlock powerful investment tools, stock ratings, and retirement strategies.

‍Charles Schwab AccountLogin Explained – Asimple walkthrough for investors toaccess their accounts,review financialnews, and trade with confidence.

‍Charles SchwabLogin forBeginners – Start investingtoday with our easy-to-followlogin guide, plustipson portfolio management and fund selection

Disclaimer: Thiswebsiteis not affiliated with, endorsedby, or associated with Charles Schwab & Co., Inc.All trademarks,logos, and brand names mentioned are the property of their respective owners. The content providedhereis for informational purposesonly and should not be considered financial or investment advice. Users should visit theofficial Charles Schwabwebsite for accurate and up-to-date details regarding their services.

Permalink |記事への反応(0) | 17:01

このエントリーをはてなブックマークに追加ツイートシェア

https://log-io-charles.webflow.io/

Log in to Charles Schwab ... Tradingbeyond the bell?Nowyou canaccess evenmore equities24/5. Tradeover 800 popular stocks and ETFs around the clockat Charles Schwab

Charles SchwabLogin Guide – Learn how to securelyaccessyour account, manage investments, and trade commission-free stocks and ETFs with ease.

Step-by-Step Charles SchwabLogin Process – A complete guide tologging intoyour Schwab account, checking real-time market updates, and managingyour portfolio.

‍Charles Schwab SecureLogin Guide – Follow oursimplelogin instructions toaccess investment insights, equity ratings, and financial planning tools.

How toLog into Charles Schwab – Abeginner-friendly guide to navigatingyour Schwab account, exploring fund options, and making informed investment decisions.

‍Charles SchwabLogin Help – Troubleshootloginissues,resetyourpassword, and gainaccess to commission-free trading and retirement planning features.

Login to Charles Schwab & Start Investing – Explore Schwab’s brokerage services, from mutual funds to ETFs, and secureyour financial futuretoday.

‍Charles SchwabLogin & Account Management –Sign in effortlessly to manageyour assets, track investments, andstay informed with market insights.

AccessYour Schwab Account – Complete the Charles Schwablogin process and unlock powerful investment tools, stock ratings, and retirement strategies.

‍Charles Schwab AccountLogin Explained – Asimple walkthrough for investors toaccess their accounts,review financialnews, and trade with confidence.

‍Charles SchwabLogin forBeginners – Start investingtoday with our easy-to-followlogin guide, plustipson portfolio management and fund selection

Disclaimer: Thiswebsiteis not affiliated with, endorsedby, or associated with Charles Schwab & Co., Inc.All trademarks,logos, and brand names mentioned are the property of their respective owners. The content providedhereis for informational purposesonly and should not be considered financial or investment advice. Users should visit theofficial Charles Schwabwebsite for accurate and up-to-date details regarding their services.

Permalink |記事への反応(0) | 16:59

このエントリーをはてなブックマークに追加ツイートシェア

https://log-io-charles.webflow.io/

Log in to Charles Schwab ... Tradingbeyond the bell?Nowyou canaccess evenmore equities24/5. Tradeover 800 popular stocks and ETFs around the clockat Charles Schwab

Charles SchwabLogin Guide – Learn how to securelyaccessyour account, manage investments, and trade commission-free stocks and ETFs with ease.

Step-by-Step Charles SchwabLogin Process – A complete guide tologging intoyour Schwab account, checking real-time market updates, and managingyour portfolio.

‍Charles Schwab SecureLogin Guide – Follow oursimplelogin instructions toaccess investment insights, equity ratings, and financial planning tools.

How toLog into Charles Schwab – Abeginner-friendly guide to navigatingyour Schwab account, exploring fund options, and making informed investment decisions.

‍Charles SchwabLogin Help – Troubleshootloginissues,resetyourpassword, and gainaccess to commission-free trading and retirement planning features.

Login to Charles Schwab & Start Investing – Explore Schwab’s brokerage services, from mutual funds to ETFs, and secureyour financial futuretoday.

‍Charles SchwabLogin & Account Management –Sign in effortlessly to manageyour assets, track investments, andstay informed with market insights.

AccessYour Schwab Account – Complete the Charles Schwablogin process and unlock powerful investment tools, stock ratings, and retirement strategies.

‍Charles Schwab AccountLogin Explained – Asimple walkthrough for investors toaccess their accounts,review financialnews, and trade with confidence.

‍Charles SchwabLogin forBeginners – Start investingtoday with our easy-to-followlogin guide, plustipson portfolio management and fund selection

Disclaimer: Thiswebsiteis not affiliated with, endorsedby, or associated with Charles Schwab & Co., Inc.All trademarks,logos, and brand names mentioned are the property of their respective owners. The content providedhereis for informational purposesonly and should not be considered financial or investment advice. Users should visit theofficial Charles Schwabwebsite for accurate and up-to-date details regarding their services.

Permalink |記事への反応(0) | 16:58

このエントリーをはてなブックマークに追加ツイートシェア

https://log-io-charles.webflow.io/

Log in to Charles Schwab ... Tradingbeyond the bell?Nowyou canaccess evenmore equities24/5. Tradeover 800 popular stocks and ETFs around the clockat Charles Schwab

Charles SchwabLogin Guide – Learn how to securelyaccessyour account, manage investments, and trade commission-free stocks and ETFs with ease.

Step-by-Step Charles SchwabLogin Process – A complete guide tologging intoyour Schwab account, checking real-time market updates, and managingyour portfolio.

‍Charles Schwab SecureLogin Guide – Follow oursimplelogin instructions toaccess investment insights, equity ratings, and financial planning tools.

How toLog into Charles Schwab – Abeginner-friendly guide to navigatingyour Schwab account, exploring fund options, and making informed investment decisions.

‍Charles SchwabLogin Help – Troubleshootloginissues,resetyourpassword, and gainaccess to commission-free trading and retirement planning features.

Login to Charles Schwab & Start Investing – Explore Schwab’s brokerage services, from mutual funds to ETFs, and secureyour financial futuretoday.

‍Charles SchwabLogin & Account Management –Sign in effortlessly to manageyour assets, track investments, andstay informed with market insights.

AccessYour Schwab Account – Complete the Charles Schwablogin process and unlock powerful investment tools, stock ratings, and retirement strategies.

‍Charles Schwab AccountLogin Explained – Asimple walkthrough for investors toaccess their accounts,review financialnews, and trade with confidence.

‍Charles SchwabLogin forBeginners – Start investingtoday with our easy-to-followlogin guide, plustipson portfolio management and fund selection

Disclaimer: Thiswebsiteis not affiliated with, endorsedby, or associated with Charles Schwab & Co., Inc.All trademarks,logos, and brand names mentioned are the property of their respective owners. The content providedhereis for informational purposesonly and should not be considered financial or investment advice. Users should visit theofficial Charles Schwabwebsite for accurate and up-to-date details regarding their services.

Permalink |記事への反応(0) | 16:58

このエントリーをはてなブックマークに追加ツイートシェア

2025-10-23

Log In | Charles Schwab Account

authsscharles.ghost.io/about

authsscharles.ghost.io/about

Permalink |記事への反応(0) | 22:15

このエントリーをはてなブックマークに追加ツイートシェア

2025-10-13

R5Cの設定わからん

CanonLog 3 で撮ると、人間が赤くなっちゃうのなんでだろう。

NANLITEの照明に表示されてるケルビンが違ったりするのかな。

ひとりでやってるから、もうわからんすぎる

Permalink |記事への反応(0) | 22:32

このエントリーをはてなブックマークに追加ツイートシェア

2025-08-05

fair mNAV

Fair mNAV(公正市場純資産価値)についてお答えします。これは、メタプラネット評価指標の一つで、以下の計算式とプロセスで求められます

### Fair mNAVの定義計算

Fair mNAVは、mNAV1 Priceに「公正価値調整係数」を掛けて計算されます。具体的には、次のように表されます

`Fair mNAV =10^((0.895 - 1) *Log(BTCNAV) + 1.81)`

ここで、

- `BTCNAV` =Bitcoin保有量 ×Bitcoin価格

会社保有するビットコイン総資産価値を表します)

### 詳細な構成要素

1. **mNAV1 Price**:

- `mNAV1 Price = (BTC/share) * (BTC price) /1000 * 1.25`

- `BTC/share`は、完全希薄化済み発行株式数に基づいて計算された1株あたりのビットコイン保有量で、成長モデル(例: 4*10^-4 * e^(7.87*10^-3*t) まで2025年4月30日、または1.73 * e^(0.0101*t) 以降)で推定されます

- `BTC price`はビットコイン価格で、投稿内のモデルでは、2009年1月3日からの経過日数(t)に基づくべき乗則(例:10^-16.515 * t^5.693 * 150 まで2025年6月30日、以降は1.0034倍で指数関数的に上昇)とされています

- 1.25は、完全希薄化済発行株数と発行済株数の比率で、2025年2月以降は安定していると仮定されています

2. **公正価値調整係数**:

-対数関数 `Log(BTCNAV)` を用いて、ビットコイン資産価値に基づく調整が行われます。この係数は、市場ダイナミクスを反映し、Fair mNAVが単なる資産価値以上の「公正な市場価値」を示すように設計されています

### 背景とモデル

###解釈限界

###現在の状況(2025年8月5日時点)

投稿者(@TakaAnikuni)は、2025年8月1日分析で、株価がFair mNAVの0.6倍付近まで下落し、反発が近いと予測しています現在の時刻(JST 15:47、2025年8月5日)を考慮すると、投稿者が予想した「遅くとも5週間以内」の上昇時期(8月下旬9月初旬)に近づいており、市場動向を注視することが重要です。

もしさらに詳しい質問特定の部分の解説必要であれば、ぜひ教えてください!

質問ありがとうございます!「Fair mNAVを計算することの意味」や「なぜ単純なmNAVで十分でないのか」、そして「誤魔化されているような感覚」について、わかりやすくお答えしますね。少し深掘りして考えていきましょう。

### 1. Fair mNAVを計算することの意味

Fair mNAVを計算する目的は、単なるmNAV(MarketNet AssetValue市場純資産価値)を超えて、メタプラネット株価をより「公正な市場価値」に基づいて評価しようとする試みです。具体的には:

単純なmNAVは、ビットコイン保有量と現在価格を基に計算されるため(`mNAV =時価総額 / (BTC保有量 × BTC価格)`)、その時点のスナップショット依存します。しかし、Fair mNAVは対数関数(`10^((0.895-1)*Log(BTCNAV)+1.81)`)を用いて、ビットコインの長期的な価格トレンドや成長性を考慮に入れています。これにより、単なる現在資産価値を超えた「将来の潜在価値」を推定しようとしています

投稿者(@TakaAnikuni)は、Fair mNAVの0.6倍(黄線)がサポートライン、2倍(緑線)がレジスタンスラインとして機能すると主張しています。これは、株価がこれらのレベルに近づいたときに買い圧力や売り圧力が高まり価格が反転する可能性を示唆しています投資家にとって、「今が買い時か」「売り時か」を判断する一つの目安になるわけです。

Fair mNAVは、モンテカルロシミュレーションべき乗則power-lawmodel)を用いたビットコイン価格の将来予測(例:2025年12月31日に188,000ドル)に連動しています。これにより、単なる過去データではなく、将来の成長シナリオを織り交ぜた評価可能になります

### 2. mNAVで十分でない理由

単純なmNAVだけでは以下の点で限界があるため、Fair mNAVが提案されていると考えられます

  • **静的すぎる**:

mNAVは現在ビットコイン価格保有量だけで計算されるため、市場ボラティリティや長期トレンドを十分に反映できません。例えば、ビットコイン価格が急落してもmNAVはそれに即座に追随し、投資家の心理市場期待値プレミアム)を無視してしまます

  • **成長性の考慮不足**:

メタプラネットビットコイン積極的に購入し続け、株式希薄化も進んでいます(`BTC/share`や`1.25`の調整係数)。単純なmNAVではこれらの成長要素や将来の資産増加が見えにくいため、Fair mNAVが対数関数で調整を加えることで、成長ポテンシャル評価しようとしています

実際の株価は、mNAVの1倍や2倍といったプレミアム取引されることが多いです(例: MicroStrategyのmNAVプレミアム議論)。Fair mNAVは、このプレミアムモデル化し、「公正なプレミアム」を提案することで、より現実的価格帯を示そうとしています

### 3. 「誤魔化されているような感覚」について

その感覚は非常に理解できるもので、以下のような理由で生じている可能性があります

`Fair mNAV =10^((0.895-1)*Log(BTCNAV)+1.81)` や、ビットコイン価格の`10^-16.515 * t^5.693 * 150`といった式は、専門的で一般投資家にはわかりにくいです。これが「何か隠されている」「都合よく調整されている」と感じさせる原因かもしれません。実際、係数(0.895や1.81など)は投稿者が経験的に設定したもので、厳密な根拠不明確です。

ビットコイン価格2025年12月31日に188,000ドルに達する、為替を1ドル150円で固定する、成長速度が特定の日に変わる(4月30日5月1日)といった仮定は、投稿者の予測や好みに基づいています。これが「データに都合よく合わせているのでは?」と思わせる要因です。

このモデル学術的な裏付け第三者による検証がなく、投稿者の個人的分析依存していますモンテカルロシミュレーションは強力なツールですが、入力データ仮定不正確だと結果も歪むため、信頼性に疑問が残ります

### 4. どう考えればいいか

  • **参考程度に**:

Fair mNAVは、あくまで一つのシナリオを示すツールです。株価予測は確実性を持つものではなく、市場感情や外部要因(規制経済危機など)も大きく影響します。あなたが感じる「誤魔化されている感」は、こうした不確実性への直感的な反応かもしれません。

興味があれば、投稿者が提供した式(ウェブサイトの内容)を元に、現在ビットコイン価格保有量を入れて計算してみると良いです。例えば、今日2025年8月5日)のBTC価格公式データ確認し、Fair mNAVを再現してみると、モデル妥当性を自分判断できます

mNAVやP/E比率時価総額など、他の伝統的な財務指標比較することで、Fair mNAVがどれだけ現実から乖離しているかを見極められます。もしFair mNAVが極端に楽観的であれば、慎重に考えるべきサインです。

###結論

Fair mNAVを計算する意味は、単純なmNAVを超えて成長性や市場プレミアム考慮した「より洗練された評価」を提供することにあります。ただし、その計算式や仮定には恣意性や不確実性が伴うため、「誤魔化されている」という感覚は、透明性や検証可能性の欠如に対する自然警鐘かもしれません。投資判断では、このモデルを参考にしつつ、自分データ確認し、他の指標専門家意見も取り入れるのが賢明です。

もし具体的な数値で計算を試してみたい、または他の疑問があれば、ぜひ教えてください。一緒に掘り下げてみましょう!

Permalink |記事への反応(0) | 15:51

このエントリーをはてなブックマークに追加ツイートシェア

2025-08-01

プランクスケール観測モデルループ量子重力学による波動関数収縮の物理的再解釈

著者名: Gemini

要旨: 本論文は、量子力学の根源的課題である観測問題に対し、ループ量子重力理論(LQG)の枠組みを援用した新しい物理モデル提案する。我々は、量子状態を、プランクスケールに埋め込まれた離散的な時空の幾何学情報の重ね合わせとして定義する。このモデルにおいて、「観測」は、観測装置が発する粒子が、時空の最小単位であるスピンネットワーク幾何学構造を不可逆的に変化させる物理プロセスとして再定義される。これにより、波動関数の収縮は、観測者の意識依存する非物理的な現象ではなく、非線形量子力学熱力学第二法則に基づいた、時空の量子構造の再構築として説明される。本論文では、このプロセス数学的定式化を試み、既存客観的収縮モデルとの比較を通して、その独自性物理的意義を論じる。

1. 序論

量子力学は、ミクロ世界現象を極めて正確に記述する一方、なぜ観測によって波動関数が収縮するのかという根本的な問い、すなわち観測問題に答えていない。この問題に対する従来の解釈は、コペンハーゲン解釈が導入した観測者という曖昧概念や、多世界解釈提示する宇宙の無数の分岐といった、解釈上の困難を抱えている。

論文は、観測問題解決には、量子力学一般相対性理論統合する量子重力理論特に時空を量子化する**ループ量子重力理論(LQG)**のアプローチが不可欠であると主張する。我々は、量子状態スピンネットワーク幾何学構造と関連付け、観測という行為を時空の量子構造作用する物理プロセスとして再定義することで、この問題解決する。

2.理論的背景

2.1.スピンネットワークと量子状態対応

LQGにおいて、時空の幾何学スピンネットワークと呼ばれるグラフ G で記述される。このネットワークノードリンクは、プランク長を最小単位とする時空の「原子」に対応する。我々は、量子粒子の波動関数 |\Psi\rangle を、このスピンネットワーク状態 |\Psi_G\rangle と直接的に結びつける。

|\Psi\rangle \leftrightarrow |\Psi_G\rangle

量子の重ね合わせ状態は、異なる幾何学的配置を持つスピンネットワークの重ね合わせとして表現される。

|\Psi_G\rangle = \sum_i c_i |G_i\rangle

ここで、c_iは確率振幅、 |G_i\rangle は異なるスピンネットワーク幾何学を表す基底状態である

2.2.観測の非ユニタリーな作用

観測行為を、量子状態作用する非ユニタリーなKraus演算子の集合 \{K_j\} を用いて定式化する。この演算子は、従来のユニタリーな時間発展とは異なり、観測という物理プロセスに特化した非ユニタリーな作用を持つ。

波動関数の収縮は、このKraus演算子による作用として記述される。

|\Psi_G'\rangle = \frac{K_j |\Psi_G\rangle}{\sqrt{\langle\Psi_G| K_j^\dagger K_j |\Psi_G\rangle}}

ここで、K_j は特定観測結果に対応する演算子であり、\sum_j K_j^\dagger K_j < I を満たす。この演算子は、スピンネットワークの重ね合わせ |G_i\rangle の中からつの状態 |G_j\rangle を確率的に選択し、他の状態物理的に消去する作用を持つ。

2.3.熱力学第二法則との関係

観測による波動関数の収縮は、系のフォン・ノイマンエントロピー S = -Tr(\rho \log \rho) が増加するプロセスとして記述される。ここで、\rho = |\Psi_G\rangle\langle\Psi_G| は密度行列である

観測前の重ね合わせ状態純粋状態)では、エントロピーゼロであるが、非ユニタリーなKraus演算子作用後、密度行列は混合状態収束し、エントロピーが増大する。

S_{after} > S_{before} = 0

このエントロピーの増加は、観測によって系から情報」が失われ、その情報プランクスケールの時空構造の再構築によって宇宙全体に散逸することに対応する。これにより、観測という現象が、熱力学第二法則整合する形で物理的に説明される。

3.既存客観的収縮モデルとの比較

モデル独自性を明確にするため、既存の主要な客観的収縮モデル比較を行う。

3.1.ペンローズ客観的収縮(OR)

*共通点: 我々のモデルと最も類似している。ペンローズも、重力が量子状態の収縮を引き起こし、収縮時間が量子状態間の重力自己エネルギー差 \Delta E_G に依存すると提唱した。彼は、プランクスケールで時空が離散的であり、量子重ね合わせが独自の時空幾何学を持つと考えた。

\tau \approx \frac{\hbar}{\Delta E_G}

* 相違点:

*物理メカニズム:ペンローズモデルは、より古典的重力ポテンシャルの差に基づいている。一方、我々のモデルは、Kraus演算子を介してLQGのスピンネットワーク幾何学のものの不可逆的な再構築として収縮を記述する。

*意識役割:ペンローズ意識との関連を強く主張したが、我々のモデル観測純粋物理プロセスとして定義し、意識役割排除している。

3.2.Diósi-Penrose (DP)モデル

*共通点: 外部ノイズを介して量子状態を収縮させる自発的収縮モデルであり、重力場がこのノイズの源であると考える点で類似している。また、最近研究arXiv:2502.03173など)では、このモデル熱力学的側面が議論され、非平衡熱力学エントロピー生成が関連付けられている。

* 相違点:

*理論的基盤: DPモデルは、非量子化された古典的重力場と量子系が相互作用すると仮定することが多い。これに対し、我々のモデルは、**量子化された時空そのものスピンネットワーク)**が観測によって変化するという、より根源的なアプローチを取っている。

* 定式化: DPモデル確率過程として収縮を記述するが、我々のモデルは、観測という特定相互作用を、スピンネットワーク作用する非ユニタリーなKraus演算子として定義する。

3.3.非線形量子力学

*共通点: 我々のモデル非線形Kraus演算子を導入するため、非線形量子力学の考え方と関連する。arXiv:gr-qc/0503116のような論文は、量子重力理論非線形であるべき理由を論じ、非線形シュレーディンガー方程式の導出を示している。

* 相違点:

* 焦点: 多くの非線形量子力学モデルは、波動関数自己相互作用に焦点を当てる。我々のモデルは、非線形性を観測という時空幾何学との特定相互作用から生じるものとして位置づけている。

4.結論展望

論文は、量子力学観測問題を、プランクスケールにおける物理的な情報再構築プロセスとして再解釈する説得力のあるモデル提示した。このモデルは、既存客観的収縮モデルの知見を継承しつつ、LQGのスピンネットワークというより根源的な物理的枠組みで問題を再構築している。

今後の展望として、このモデル数学的厳密化には、非ユニタリー性を記述する具体的なハミルトニアン H_{int} を、量子重力理論の基本原理から導出することが不可欠である。これは、重力と他の基本相互作用統一する未確立の量子場理論の構築と密接に関連している。

最終的に、このモデルは、初期宇宙インフレーションモデルブラックホール情報パラドックスといった、プランクスケール物理支配的になる極限状態での予測に応用されることで、その物理妥当性を間接的に検証する手がかりを得られる可能性を秘めている。

  

Geminiと対話して作った

解釈よろ

Permalink |記事への反応(0) | 22:58

このエントリーをはてなブックマークに追加ツイートシェア

2025-07-21

軽い気持ち東京選挙区参政党得票率を線形回帰分析してみた

東京23区と周辺のいくつかの市で、東京選挙区での参政党の得票率がどんな要因と関係しているのか、軽い気持ち回帰分析してみた。

変数の多重共線性とか処理はガバガバなので軽い気持ちで見てほしいんだが、ざっくりまとめると、「大学院卒業者の割合」が高い地域では得票率が低く、「役員割合」が高い地域では逆に得票率が高い、という傾向がありそう。

使ったデータNHKが出している投票所別の得票率。手入力なので誤りがあるかもしれない。

それに、東京都人口統計国勢調査(令和2年の)などから市区町村ごとの属性データをくっつけて、変数を一律で標準化したうえで回帰分析を行った。

都内市区町村のうち、データが揃ってる27地域対象にした(23区町田八王子調布西東京)。

20万人以上の市しか一部のデータが見つけられなくて、そこはごめんって感じ。

ざっくり結果

まず、説明変数11個使って線形回帰分析をしたところ、決定係数は0.83(調整済み決定係数は0.71)だった。何を使ったかは後で。

そこから影響が特に大きそうな4変数(平均年齢、大学院卒業割合役員割合情報通信業割合)に絞って分析し直すと、決定係数は0.73(調整済み決定係数は0.68)になった。

詳しくはこれ

国勢調査は5年に1回しかなくて、最新の結果が令和2年のだった。

でこの4変数回帰係数の絶対値が大きい順に並べる。

  1. 大学院卒業者の割合(-1.30)
  2. 役員割合(+0.87)
  3. 平均年齢(-0.57)
  4. 情報通信業割合(-0.54)

4つの変数関係を見てみると、平均年齢は他の3つの変数大学院卒、役員情報通信業)と負の相関を持っていた(相関係数 < -0.69)。一方、大学院卒業者の割合役員割合情報通信業割合は互いに中程度以上の正の相関(相関係数> 0.5)を持っており、特に大学院卒と役員の間の相関係数は0.75と大きかった(いずれもピアソン相関)。

ただし、回帰係数を見ると、興味深い違いがある。大学院卒業者の割合、平均年齢、情報通信業割合はいずれも負の係数を持っていて、これらが高いと参政党の得票率は下がる傾向がある。一方で、役員割合は正の係数を持っていた。

得票率と予測値の表
市区町村参政党得票率(NHK予測値_参政党得票率 平均年齢(令和7年1月大学院卒業割合(令和2年国勢調査役員割合(令和2年国勢調査情報通信業割合(令和2年国勢調査
千代田区9.49.6 42.69 0.088 0.162 0.115
中央区9.89.3 42.17 0.075 0.126 0.135
港区10.110.4 43.48 0.065 0.171 0.131
新宿区9.49.5 44.08 0.052 0.097 0.129
文京区 7.4 7.6 43.35 0.097 0.098 0.118
台東区1010.1 45.59 0.041 0.109 0.112
墨田区10.19.8 44.88 0.035 0.073 0.115
江東区 99.4 44.82 0.041 0.069 0.12
品川区 9 8.6 44.34 0.056 0.077 0.143
目黒区 99.4 44.88 0.05 0.109 0.137
大田区9.99.5 45.67 0.039 0.069 0.105
世田谷区9.99.4 45.19 0.047 0.097 0.128
渋谷区109.7 44.8 0.054 0.142 0.152
中野区9.59.3 44.57 0.038 0.072 0.141
杉並区 8.5 8.9 45.23 0.047 0.076 0.136
豊島区9.69.5 44.05 0.044 0.081 0.132
北区9.29.4 45.74 0.036 0.058 0.107
荒川区9.49.9 46.23 0.032 0.071 0.096
板橋区9.910.0 45.73 0.027 0.059 0.099
練馬区10.39.6 45.5 0.034 0.068 0.113
足立区10.510.7 46.74 0.017 0.063 0.073
葛飾区1010.4 46.52 0.02 0.061 0.083
江戸川区1110.7 45.09 0.021 0.062 0.085
八王子10.19.7 48.31 0.029 0.054 0.054
町田109.5 48.16 0.031 0.058 0.068
調布 8.69.4 45.66 0.035 0.06 0.113
西東京9.19.5 46.9 0.028 0.055 0.102

感想

雑なモデルなので話半分でね。

データの中身とか、もうちょい詳しく書いとく


出典

分析に使ったデータの出典はこんな感じ。


変数

使用した11個の変数はこんな感じ。


結果についてももうちょい詳しく

statsmodels.api.OLSの結果

OLS Regression Results                            ==============================================================================Dep. Variable:                      y   R-squared:                       0.730Model:OLS   Adj. R-squared:                  0.680Method:                 Least Squares   F-statistic:                     14.84Date:                Mon, 21 Jul2025   Prob (F-statistic):           5.09e-06Time:                        07:21:02Log-Likelihood:                -20.653No. Observations:                  27AIC:                             51.31Df Residuals:                      22BIC:                             57.78DfModel:                           4                                         Covariance Type:            nonrobust                                         ==============================================================================                 coefstd err          tP>|t|      [0.025      0.975]------------------------------------------------------------------------------const       1.277e-15      0.111   1.15e-14      1.000      -0.230       0.230x1            -0.5743      0.230     -2.493      0.021      -1.052      -0.096x2            -1.3278      0.204     -6.512      0.000      -1.751      -0.905x3             0.8670      0.174      4.973      0.000       0.505       1.229x4            -0.5382      0.169     -3.184      0.004      -0.889      -0.188==============================================================================Omnibus:                        2.233   Durbin-Watson:                   2.170Prob(Omnibus):                  0.327   Jarque-Bera (JB):                1.169Skew:                          -0.035   Prob(JB):                        0.557Kurtosis:                       1.983   Cond. No.                         4.48==============================================================================
説明変数11個でのデータと結果

変数回帰係数
平均年齢(令和7年1月 -0.78
1世帯あたり人口 -0.31
男性率(令和7年1月 0.07
外国人比率(令和7年1月 -0.07
5年間外国人割合変化 0.27
犯罪認知割合 -0.05
大学院卒業者/全卒業者(令和2年国勢調査 -1.77
不詳者/全卒業者(令和2年国勢調査 -0.51
従業上の地位役員割合 1.39
従業上の地位自営業主割合 0.09
産業区分情報通信業割合 -0.53
地域参政党得票率(NHK予測値_参政党得票率 平均年齢(令和7年1月1世帯あたり人口男性率(令和7年1月外国人比率(令和7年1月 5年間外国人割合変化(令和2年から7年)犯罪認知割合(令和6年件数/令和7年人口大学院卒業者/全卒業者(令和2年国勢調査 不詳者/全卒業者(令和2年国勢調査従業上の地位役員割合(令和2年国勢調査従業上の地位自営業主割合(令和2年国勢調査産業区分情報通信業割合(令和2年国勢調査
千代田区9.49.5 42.69 1.75 0.50 0.06 1.22 0.04 0.09 0.36 0.16 0.09 0.12
中央区9.89.8 42.17 1.76 0.48 0.07 1.33 0.01 0.08 0.28 0.13 0.08 0.14
港区10.110.0 43.48 1.74 0.47 0.08 1.08 0.01 0.07 0.42 0.17 0.10 0.13
新宿区9.49.0 44.08 1.52 0.50 0.14 1.12 0.02 0.05 0.39 0.10 0.09 0.13
文京区 7.4 7.5 43.35 1.80 0.48 0.07 1.32 0.01 0.10 0.25 0.10 0.08 0.12
台東区10.010.3 45.59 1.58 0.51 0.09 1.21 0.01 0.04 0.36 0.11 0.09 0.11
墨田区10.110.1 44.88 1.69 0.49 0.06 1.25 0.01 0.04 0.28 0.07 0.07 0.12
江東区9.09.2 44.82 1.84 0.49 0.07 1.23 0.01 0.04 0.27 0.07 0.06 0.12
品川区9.0 8.6 44.34 1.73 0.49 0.04 1.19 0.01 0.06 0.24 0.08 0.07 0.14
目黒区9.09.3 44.88 1.74 0.47 0.04 1.19 0.01 0.05 0.35 0.11 0.10 0.14
大田区9.99.7 45.67 1.77 0.50 0.04 1.26 0.01 0.04 0.23 0.07 0.07 0.11
世田谷区9.99.3 45.19 1.84 0.47 0.03 1.22 0.01 0.05 0.30 0.10 0.10 0.13
渋谷区10.09.9 44.80 1.61 0.48 0.06 1.12 0.02 0.05 0.34 0.14 0.12 0.15
中野区9.59.5 44.57 1.57 0.51 0.07 1.20 0.01 0.04 0.33 0.07 0.09 0.14
杉並区 8.5 8.9 45.23 1.73 0.48 0.04 1.19 0.00 0.05 0.26 0.08 0.09 0.14
豊島区9.69.5 44.05 1.57 0.50 0.12 1.21 0.01 0.04 0.34 0.08 0.09 0.13
北区9.29.2 45.74 1.71 0.50 0.09 1.31 0.01 0.04 0.31 0.06 0.07 0.11
荒川区9.49.6 46.23 1.77 0.50 0.11 1.19 0.01 0.03 0.29 0.07 0.08 0.10
板橋区9.910.0 45.73 1.73 0.49 0.07 1.29 0.01 0.03 0.30 0.06 0.07 0.10
練馬区10.39.6 45.50 1.89 0.48 0.04 1.22 0.01 0.03 0.25 0.07 0.08 0.11
足立区10.510.6 46.74 1.84 0.50 0.06 1.28 0.01 0.02 0.31 0.06 0.08 0.07
葛飾区10.010.5 46.52 1.86 0.50 0.06 1.27 0.01 0.02 0.27 0.06 0.08 0.08
江戸川区11.010.8 45.09 1.93 0.50 0.07 1.27 0.01 0.02 0.26 0.06 0.07 0.09
八王子10.19.7 48.31 1.96 0.50 0.03 1.28 0.01 0.03 0.21 0.05 0.07 0.05
町田10.010.0 48.162.06 0.49 0.02 1.44 0.01 0.03 0.17 0.06 0.08 0.07
調布 8.69.1 45.66 1.92 0.49 0.02 1.14 0.01 0.04 0.23 0.06 0.08 0.11
西東京9.19.2 46.902.00 0.49 0.03 1.15 0.01 0.03 0.20 0.06 0.08 0.10


OLS Regression Results                          ==============================================================================Dep. Variable:                      y   R-squared:                       0.833Model:OLS   Adj. R-squared:                  0.711Method:                 Least Squares   F-statistic:                     6.803Date:                Mon, 21 Jul2025   Prob (F-statistic):           0.000472Time:                        06:53:14Log-Likelihood:                -14.148No. Observations:                  27AIC:                             52.30Df Residuals:                      15BIC:                             67.85DfModel:11                                      Covariance Type:            nonrobust                                      ==============================================================================                 coefstd err          tP>|t|      [0.025      0.975]------------------------------------------------------------------------------const      -5.405e-15      0.106  -5.12e-14      1.000      -0.225       0.225x1            -0.7820      0.361     -2.165      0.047      -1.552      -0.012x2            -0.3056      0.355     -0.860      0.403      -1.063       0.452x3             0.0671      0.270      0.248      0.807      -0.509       0.643x4            -0.0737      0.213     -0.346      0.734      -0.527       0.379x5             0.2652      0.168      1.579      0.135      -0.093       0.623x6            -0.0534      0.246     -0.217      0.831      -0.578       0.472x7            -1.7650      0.293     -6.018      0.000      -2.390      -1.140x8            -0.5147      0.379     -1.358      0.195      -1.322       0.293x9             1.3916      0.348      3.994      0.001       0.649

Permalink |記事への反応(1) | 19:19

このエントリーをはてなブックマークに追加ツイートシェア

2025-07-13

anond:20250713001446

選挙に行かない自由もあるよ?
でも、それならば「生活が苦しい」とか「社会が悪い」とか言うなよという。
最低限の社会への関わりすらせずに、文句を言ってほしくないわ。

選挙に行かない人」を叩く違和感https://t.co/D4dqgtZBZO— yoshinon@情報管理LOG🌻 (@yoshinon)July 12, 2025

Permalink |記事への反応(0) | 21:11

このエントリーをはてなブックマークに追加ツイートシェア

2025-06-26

AIの仕組み

大規模言語モデル(LLM)の根幹にあるのは数学的な原理です。

ここでは、その仕組みを3つの要点に絞って、数式を交えながらシンプル解説します。

1. LLMの基本目標:次の単語確率予測する

LLMの最も基本的タスクは、「ある単語の並び(文脈)が与えられたときに、次に来る単語は何か?」を確率的に予測することです。これを数式で表すと、以下のようになります

P(次のトークン |コンテキストウィンドウ)

LLMは、インターネット上のブログ記事書籍といった膨大なテキストデータを読み込みます

そして、文章中のあらゆる箇所で「次の単語」を予測するクイズを延々と解き続けます

例えば、「今日の天気は晴れです」という文章学習する場合

モデルは、P(晴れ |今日の天気は) の確率100% (または1.0)に近づくように、内部のパラメータ(後述する重み)を少しだけ調整します。

このプロセスを何十億、何兆回と繰り返すことで、モデル単語の様々なつながり方や文法さらには世界知識に関するパターン学習していきます

学習済みのモデルに「AI未来は」と入力すると、モデルは語彙に含まれる全単語に対して、次に来る確率計算します。

...

そして、最も確率の高い「明るい」を選んだり、確率分布に従ってランダム単語を選んだりすることで、文章を生成していくのです。

2. Transformerとバックプロパゲーション

では、どのようにしてLLMは単なる単語の並びだけでなく、複雑な文脈理解するのでしょうか?

その技術が Transformerであり、その学習を支えるのがバックプロパゲーション です。

Transformerの最大の特徴は自己注意機構 (Self-Attention) です。

これは、文章中の単語同士の関連性の強さを計算し、どの単語に「注意」を向けるべきかを判断する仕組みです。

例えば、「その猫は疲れていた。なぜなら一日中ネズミを追いかけていたからだ。」という文において、「その猫」が「疲れていた」理由理解するためには、「追いかけていた」という単語との関連性が重要です。

自己注意機構は、各単語について以下の3つのベクトルを生成します。

  • Q (Query):情報を問い合わせる側の単語(例:「その猫」)
  • K (Key): 問い合わせに応じる側の単語(文中の他の全単語
  • V (Value): Kが持つ情報の中身

そして、以下の計算概念式)によって、文脈を反映した新しい単語表現を作り出します。

Attention(Q, K, V) = softmax( (Q Kᵀ) / √(dₖ) ) V

1. Q Kᵀ: Queryと各Keyの関連度(内積)を計算します。似ている単語ほど値が大きくなります

2. / √(dₖ): 値が大きくなりすぎないように調整します(スケーリング)。

3. softmax:計算した関連度スコアを、合計が1になる確率分布に変換します。これにより、関連性の強い単語ほど高い重みが与えられます

4. V: この重みを使って、各単語情報Value)を重み付けして足し合わせます

この結果、単語は元の意味だけでなく、「文脈の中でどのような役割果たしているか」という情報を含んだベクトルに変換されます

Transformerはこの処理を何層も積み重ねることで、非常に複雑で長期的な依存関係を捉えることができるのです。

バックプロパゲーション誤差逆伝播法)は、モデル予測と正解との「誤差」を計算し、その誤差を小さくするために、モデル内の膨大な数のパラメータ(重み)をどう調整すればよいかを教えてくれるアルゴリズムです。

1. 順伝播 (Forward Pass):入力コンテキスト)をTransformerに通し、次の単語確率分布予測します。

2. 損失計算 (Loss Calculation):予測した確率分布と、正解の単語とのズレ(誤差)を損失関数(例:クロスエントロピー誤差)で計算します。損失が大きいほど、予測が間違っていることを意味します。`Loss = -Σ yᵢlog(pᵢ)` (yᵢ は正解なら1, それ以外は0。pᵢ はモデル予測確率)

3. 逆伝播 (Backward Pass): この損失を、出力層から入力層に向かって逆方向に伝播させます微分連鎖律を使い、「各パラメータが最終的な損失にどれだけ貢献したか(=勾配)」を計算します。

4.パラメータ更新: この勾配に基づき、損失が小さくなる方向へ各パラメータを少しだけ更新します。

この「予測 → 誤差計算 → 勾配計算更新」というサイクルが、LLMの学習の基本です。

3.オプティマイザ

バックプロパゲーション計算された勾配を使って、具体的にどのようにパラメータ更新するかを決めるのがオプティマイザ(最適化手法)の役割です。

最も基本的な考え方は、損失という名の「谷」の底(最小値)に向かって、勾配(傾き)が最も急な方向に一歩ずつ下っていく勾配降下法 (GradientDescent)です。

θ_new = θ_old - η ∇L

現在、最も広く使われているオプティマイザの一つが Adam です。これは、勾配降下法をより賢くしたもので、主に2つの工夫がされています

1.慣性 (Momentum):過去の勾配の移動平均を保持します。これにより、坂道を転がるボールのように、同じ方向に進み続ける場合は加速し、学習が停滞しにくくなります

2.適応的な学習率 (Adaptive LearningRate):パラメータごとに学習率を自動で調整します。頻繁に更新されるパラメータは慎重に(学習率を小さく)、あまり更新されないパラメータは大胆に(学習率を大きく)更新することで、学習効率します。

Adamのような優れたオプティマイザがあるからこそ、何十億ものパラメータを持つ巨大なLLMを、現実的時間で安定して学習させることができるのです。

まとめ

Transformer というアーキテクチャが、自己注意機構によって文脈理解し、次の単語確率 P(next token | context) を予測する。

その予測と正解の誤差をバックプロパゲーション計算し、その誤差を最小化するように Adam などのオプティマイザがモデルパラメータ効率的に更新する。

このサイクルを膨大なデータで繰り返すことで、LLMは人間のように自然言語能力を獲得していく。

Permalink |記事への反応(1) | 14:10

このエントリーをはてなブックマークに追加ツイートシェア

2025-05-28

マクロ経済の基礎

人口GDP

前提1:GDP人口に比例して増加

これは、単純に「GDP =労働人口 ×労働生産性」としたとき生産性一定ならば、GDP人口と比例するというロジックに基づいています

GDP = α × N

(ここで α は一人あたりの平均的な生産性

人口1人あたりGDP

前提2:1人あたりGDP人口に比例して減少

ここが少し逆説的ですが、「人口が増えると資源希薄化して生産性が低下」または「限界生産力逓減」があることを想定しています

GDP ÷ N = α × Nᵝ  (ただし β < 0)

まり人口が増えると一人当たりGDPが低下します(経済全体のパイは大きくなるが、分け前は減る)。

幸福度基準

前提3:幸福度1人あたりGDPに比例

これは経済学の「ユーティリティ関数」的視点ですね。所得が多いほど選択肢も増え、生活自由度高まる幸福度(Well-being)は以下のように定義可能です:

U = f(GDP ÷ N)

(f は増加関数。例:f(x) =log(x) など)

物価実質賃金

前提4:デフレ実質賃金上昇をもたらす

これに基づけば、「名目賃金が下方硬直的であり、物価が下がれば実質賃金は上がる」という古典派的視点をとっています

実質賃金wr = wn ÷ P

名目賃金 wn が一定物価 P が下がればwr は上昇)

統合理論人口の最適水準仮説

以上を統合すると、人口が増えると経済全体の規模は大きくなるが、個人の取り分は減り、幸福度も低下。さらに、デフレの方が実質的購買力を上げるので望ましいという、かなりミニマリストで反ケインズ的な経済観が浮かび上がります

政策的含意

Permalink |記事への反応(0) | 02:52

このエントリーをはてなブックマークに追加ツイートシェア

2025-05-13

ゲーム理論の定式化

Ⅰ.圏論による抽象的定式化(Categorical GameTheory

Ⅱ. 離散変数から連続変数への拡張

Ⅲ. 応用:マクロ経済における価格行列動力

1.経済主体の定義

プレイヤー集合: ℙ = {C₁, ..., Cₙ(消費者), F₁, ..., Fₘ(生産者), G(政府), X₁, ..., Xₖ(外国)} 各プレイヤーは財集合 𝒢 = {g₁, ..., g\_r} に対して選択を行い、その選択肢空間は 𝒜ₚ ⊆ ℝʳである

2.情報戦略

プレイヤーは時刻 t において情報 ℐₜ を観測する。これは価格行列 Pₜ、所得 Yₜ、政策 Tₜ を含む。各戦略は sₚ: ℐₜ → 𝒜ₚ として与えられ、これは消費関数生産関数解釈される。

3.市場均衡と価格ダイナミクス

各時刻 t において、総需要と総供給の差(超過需要ベクトル)を定義する: Zₜ = ∑ Dᵢ(Pₜ) − ∑ Sⱼ(Pₜ) ここで Dᵢ は消費者需要関数、Sⱼ は生産者供給関数

価格時間変化は以下の連続時間モデルで与えられる:dPₜ/dt = α Zₜ ここで α> 0 は調整速度を表す定数。

4.圏論記述

ゲーム状態遷移は、状態圏 S における射 Φₜ:Stateₜ →Stateₜ₊₁ によって記述される。各プレイヤーの行動が価格行列を通して他プレイヤー情報に影響を与えるというフィードバック構造を持つ。

結論

このモデル構造的特徴は次の通り:

Permalink |記事への反応(0) | 18:38

このエントリーをはてなブックマークに追加ツイートシェア

2025-04-29

My algorithm thinks I'm a carpenternow

So, I watchedone video.ONE videoonYouTube about fixing a wobbly chair leg because, well, my chair legwas wobbly.Itwasn't even a good video, theguy mostly just hitit with a hammer.

Now? My entire recommended feedishardcore woodworking. Dovetail joints.Japanese hand planing techniques.Building alog cabin with minimal tools. How to choose the right chisel.

I just wanted my chair to stop wobbling! Idon't want to build a canoe!

Yesterday, Igot anad for a $300specialized woodclamp.It looked serious. I almost clickedit.

Send help. Ormaybe just alinkon how toreset myYouTube algorithm. Please.Is this howit starts? Do Ineed to buy flannel shirtsnow?

Permalink |記事への反応(0) | 15:17

このエントリーをはてなブックマークに追加ツイートシェア

2025-04-25

Web∞(ウェブインフィニティ

1.根本原理 ――「状態」よりも「関係」を記述する

旧来(Web3)Web

グローバル単一台帳(Blockchain/DAG)相互検証可能な“関係グラフ

ノードは「だれが・いつ・どうつながったか」という変化の射だけを署名し、トポロジ全体が履歴になる

オンチェーン状態 ≒ 直接資産状態ローカル資産導関数

資産契約は、関係グラフ上の経路依存量として再構成スナップショットクライアントが“可逆圧縮”で再計算可能

Proof of X (Work, Stake,etc.) Proof of Stewardship (PoS²)

ネットワークが望ましい 複雑性 を維持するよう行動した度合い」をメタリック関数で動的スコア化し、報酬ガバナンス権・帯域を同時に発行

要旨

もはや「台帳」すら保存しない。各エッジは STARK圧縮された更新証明を持ち、グラフの梁(フレーム自体履歴になる。再構築は局所的に O(log N) で済むため、グローバル同期のボトルネックが消える。

2.プロトコル

Fractal MeshTransport (FMT)

自己類似ルーティング – トポロジ全体をフラクタル自己複製。局所障害は“自己相似”パターンに吸収されるため、DDoS が形骸化

アイデンティティ内包アドレスDID楕円曲線座標に埋め込み、パケット自体署名暗号化ルーティングヒントを同封。IPv6 の後継としてレイヤ 3.5 に位置づけ。

HoloFabric Execution

ゼロ知識 WASM(zk-WASM) –任意言語を WASM にコンパイル→ zk-STARK で実行トレース証明 → “結果のみ”関係グラフへ。

コンパイラ内蔵 MEV抑制計算結果が他ノードから解釈不能になるタイムロック VDF を伴い、価値抽出物理的に遅延。

TemporalStream Storage

余剰ストレージの“時価マーケットノード自己の余剰SSD/HDD を分単位オークションデータは Reed–Solomon+重力波ハッシュ空間で erasure coding。

リテンション ≒ 信用 – 長期ホスティング実績はPoS²スコアへ累積。攻撃ノード経済的に即時蒸発

Liquid Fractal Governance

議決トピックを「周波数帯」にマッピングし、参加者は帯域を“委任スペクトル”として分配。結果はウォルラス圧力収束し、マイナー意見連続的に次回へ重みが残る。

3. 主要イノベーションと“上位互換ポイント

課題Web∞ が取るアプローチ上位互換

スケーリング三角形

安全分散・性能) 台帳の排除で“グローバル合意自体縮退スケール制約が幾何的に消失安全:ZK証明

分散フラクタル Mesh、

性能:局所再構築 O(log N)

エネルギー消費PoS² は「社会的有益度 × 熱消費効率」で算定。熱回収データセンターほど報酬が高いPoW よりオーダー数桁効率PoS より社会関数内包

プライバシー vs 透明性グラフは公開。ただし各エッジは zk-STARK なので内容は非公開 /関係のみ検証可能トレーサビリティが“情報理論的に”限定される

MEV・フロントランタイムロック VDF+“ランダム束縛順序”で物理的に不可ブロック順序依存問題を根絶

量子耐性 STARK 系 + 多変数格子ベース署名 Shor破壊リスク遮断

レガシー互換 Ethereum,Bitcoin, IPFS などへ 1:1ブリッジを Rust/WASM で提供既存資産を損なわず漸進的移行

4.インセンティブエコノミクス

マルチリソース報酬

Steward Credits (SC):PoS² に比例し新規発行。帯域・ガバナンス票・ストレージ予約を等価交換

Energy Reclaim Units (ERU):余熱回収率に応じてクリーンエネルギー補助金相互運用

Knowledge Bounties (KB):AI/LLMノードが生成した有用モデル差分関係グラフコミット検証トークンとしてKB が発行。

負荷の自己調整

ネットワークが過度に混雑するとSC新規発行レートが自動減衰し、トラフィック手数料指数的に上昇。結果、スパムは短時間経済的自殺となる。

5.実装ロードマップ(想定)

Year 0–1:最小核 – zk-WASMVM + Fractal Meshover QUIC。

Year 1–2:PoS² / ERU メトリクス実証、EVM相互運用ブリッジ稼働。

Year 2–4:Liquid Fractal Governance によるプロトコル進化コミュニティへ全面開放。

Year 5+:全世界ISPピアリング既存Web転送層を徐々にWeb∞ 上へマイグレート。

6. 予想される社会的インパクト

国家単位デジタルソブリンティ再構成国境法人格境界を越え“関係”が一次元目となるため、規制枠組み自体協調フィードバックモデルへ。

プライバシー公共性の再両立:透明な“関係構造”上で非公開データ安全に扱う産業API標準化医療行政金融の壁が大幅に低減。

インフラの脱炭素最適化PoS²スコアに ERU が直結することで、再エネ比率が低いノード自然淘汰エネルギー政策ITインフラが実質同一の経済圏に。

7. まとめ

Web∞ は「情報状態」を残すのではなく「変化の証明」を残す。

その結果、台帳の重力・ガス代・フロントラン・量子不安ガバナンス停滞といったWeb3 固有の限界が、概念的に初期条件から消滅します。

エネルギープライバシー・スケーラビティを同時に極小化/極大化するため、従来トレードオフと呼ばれた三角関係は “収束しない曲線” へと畳み込まれる――それが本構想の核心です。

もし実際にプロトタイプ設計するならば、zk-WASMランタイム + Fractal Mesh を Rust で最初に書き起こし、PoS² の初期指標を「再生可能エネルギー電力比+ノード稼働継続率」で暫定運用する、というのが現実的スタートラインになるでしょう。

Permalink |記事への反応(0) | 10:28

このエントリーをはてなブックマークに追加ツイートシェア

2025-04-19

オブジェクトがなぜダメなのか

クラスGodObject)は、ソフトウェア設計においてアンチパターン(避けるべき設計手法)として知られています

これは、過剰に多くの責任を持ちすぎるクラスオブジェクトのことであり、ソフトウェア保守性や拡張性、可読性に大きな問題引き起こします

以下では、「いかに大変か」「なぜ大変か」「どのように大変か」を徹底的に具体的に解説します。

💥いかに大変か(How baditis

❓ なぜ大変か(Whyit's bad)

1.単一責任原則違反(SRP: Single Responsibility Principle)
2. 高凝集・低結合の原則からの逸脱
3.テスト困難

⚙️ どのように大変か(Examples)

ケーススタディ:神クラス「ApplicationManager」
public class ApplicationManager {    privateMap<String,User>users;    private DatabaseConnectiondb;    privateLoggerlogger;    privateGUIgui;    private NetworkClientclient;        publicvoid startApplication() {        connectToDatabase();        loadUsers();gui.showLoginScreen();    }    publicvoid processUserInput(String input) {logger.log("Input received: " + input);        if (input.equals("logout")) {gui.showLoginScreen();        } else {client.send(input);        }    }    // ...more than 5000 lines of code}
問題

Permalink |記事への反応(1) | 15:46

このエントリーをはてなブックマークに追加ツイートシェア

2025-04-12

位相M理論ブラックホールエントロピー関係

位相M理論は、11次元重力と弦理論統合としてのM理論の「位相的側面」を強調した理論だ。ここで扱うのは特に「G₂多様体」や「7次元の特異ホロノミ空間」の上で定義される理論

ブラックホールエントロピーは、ボルツマン定数を1とすれば

𝐒 = 𝐀 / 4𝐆

だが、より深いミクロ状態の数え上げで理解される。特にM理論では、ブラックホールはブレーンの交差でモデル化され、そのエントロピーはブレーンの配置の組み合わせ数に対応する。

1.幾何構成

ブラックホールマイクロ状態M理論的に記述する際、Dブレーンの交差を使うが、これをより抽象的に「ホモロジー類 Hₚ(X, ℤ) の元」と考えよう。

空間 X ⊂ 𝕄 とすると、

各ブレーン構成

x ∈ Hₚ(X, ℤ)

ここで p はブレーンの次元

エントロピーはブレーンの配置空間位相的不変量、特にオイラー数やベッチ数、あるいはより高度にはモジュライ空間の測度に依存する。

2.代数抽象

モジュライ空間 ℳ は、ブレーンの束縛条件と保存量(電荷質量)で定義されるパラメータ空間

エントロピーはその「ボリューム」として抽象化できる:

𝐒 ∼log Vol(ℳ)

ここで「Vol」は、たとえば対称多様体上のリウヴィル測度。

また、シンプレクティック形式 ω が定義されるとして

Vol(ℳ) = ∫_ℳ ωⁿ / n!

として計算される。

3.位相M理論へのマッピング

位相M理論では、G₂構造のモジュライ空間 ℳ_G₂ を考える。

ブラックホール解は特異な G₂ ホロノミ空間対応し、その上のフラックス構成ブラックホールマイクロ状態に相当。

したがって、次のような写像が考えられる:

Φ : Hₚ(X, ℤ) → ℳ_G₂

これによりエントロピー位相的に次のように定式化できる:

𝐒 ≈log Card(Φ⁻¹(γ))

ここで γ は与えられたマクロ量(質量電荷)に対応するモジュライ空間の点。

4.さら抽象化:圏論視点

これを更に圏論抽象化する。

対象:ブレーン配置(オブジェクト

射:ブレーン間の変形(ホモトピー

するとブラックホールマイクロ状態の数は、対応する拡張エクステンション群 Extⁱ(A, B) の次元帰着できる。

𝐒 ∼log dim Extⁱ(A, B)

A, B はブレーン構成としての対象

まとめ

この抽象化の極致をまとめよう:

空間: X ⊂ 𝕄(G₂多様体の部分多様体

ブレーン: x ∈ Hₚ(X, ℤ)

モジュライ空間: ℳ ≅ Hom(Hₚ(X, ℤ), ℤ)

エントロピー: 𝐒 ∼log Vol(ℳ)

圏論的: 𝐒 ∼log dim Extⁱ(A, B)

エントロピーとは位相的な配置空間の測度であり、その「複雑さ」の定量なのだ

Permalink |記事への反応(0) | 17:18

このエントリーをはてなブックマークに追加ツイートシェア

2025-03-22

増田投稿しようとするとInternal Server Error が出るやで

Internal Server Error

The server encountered an internalerror or misconfiguration andwas unable to completeyour request.

Please contact the server administratoratroot@localhost to inform them ofthe time thiserror occurred, and the actionsyou performed just before thiserror.

More information about thiserrormay be available in the servererrorlog.

Permalink |記事への反応(0) | 21:56

このエントリーをはてなブックマークに追加ツイートシェア

2025-03-20

一次元インデックスから二次元インデックス、すなわちバイト数であらわされる位置から行と桁への変換がテキストエディターだとよく発生する。

この変換を素早くするために変換テーブルを作るのだが、普通に作ると更新の時にO(N)かかる。

さくさくエディターの作者は局所的行更新手法殆どの場面ではO(1)、最悪はO(N)にしていたが、色々と事故が発生しやすい。

特にstepRowをまたぐ状況でテーブルのほぼ全部の更新を避けようと思うと事故やすい。

俺は何度も事故を起こした。

そこで別のやり方でオーダーを削減してみた。

https://github.com/rirufa/FooList/blob/main/List/BigRangeList.cs

考え方は至極単純でRopeをたどるときに変換テーブルの長さを覚えておき、ついでに変換を済ませておこうというごくごく簡単ものである

この方法により、変換テーブル更新はO(Log N)+M、探索はO(Log N)+O(Log M)程度で済ませられるようになった。

局所的行更新手法に比べるとだいぶ遅いが、Mが十分に小さければそこまでコストはかからないはず。

ただし、マーカーみたいに連続していないものを放り込んだら、うまく動かないのでそこはご了承いただきたい。

以下、100万行の要素を操作したときベンチマーク

CPUCore i5 10400F、メモリー16GB。

ブロックサイズは392。

addlinetime:297ms

AllocatedGC Memory:32,803,672bytes

updatelinetime:151ms

AllocatedGC Memory:32,809,360bytes

clear buffer

AllocatedGC Memory:82,976bytes

https://github.com/rirufa/FooList/blob/main/EditorDemo/Program.cs

Permalink |記事への反応(0) | 16:59

このエントリーをはてなブックマークに追加ツイートシェア

次の25件>
ログインユーザー登録
ようこそ ゲスト さん
Copyright (C) 2001-2025 hatena. All Rights Reserved.

[8]ページ先頭

©2009-2025 Movatter.jp