-
Notifications
You must be signed in to change notification settings - Fork 0
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
0 parents
commit b2e71f2
Showing
411 changed files
with
9,950 additions
and
0 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,24 @@ | ||
# Note | ||
``` | ||
guid: B7c{<Y&(}` | ||
notetype: AI-Vocabulary-Style | ||
``` | ||
|
||
### Tags | ||
``` | ||
``` | ||
|
||
## Front | ||
AUC曲線 | ||
|
||
## Back-EN | ||
AUC curve (Area Under the Curve) | ||
|
||
## Back-FR | ||
Courbe AUC (aire sous la courbe) | ||
|
||
## Reading | ||
AUCきょくせん | ||
|
||
## Sentence | ||
ROC曲線の下の面積を AUC (Area Under the Curve) と呼び、分類モデルの評価指標として用いられる。AUC が 1 のときが最良であり、ランダムで全く無効なモデルでは 0.5 となる。 |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,29 @@ | ||
# Note | ||
``` | ||
guid: Ir7>YWDdCT | ||
notetype: AI-Vocabulary-Style | ||
``` | ||
|
||
### Tags | ||
``` | ||
``` | ||
|
||
## Front | ||
BERT言語モデル | ||
|
||
## Back-EN | ||
Bidirectional Encoder Representations from Transformers, BERT | ||
|
||
## Back-FR | ||
Représentations de l'encodeur bidirectionnelle de Transformers, Bert | ||
|
||
## Reading | ||
BERT言語モデル | ||
|
||
## Sentence | ||
BERTは、自然言語処理の事前学習用の Transformer ベースの機械学習手法である。 | ||
- Masked language model | ||
単方向制約を超えた双方向(Bidirectional)の言語モデルを構築するために、BERTでは事前学習タスク/損失関数として MLM を採用した。 | ||
MLMでは部分マスクされた系列を入力としてマスク無し系列を予測し、マスク部に対応する出力に対して一致度を計算し学習する。モデルはマスクされていない情報(周囲の文脈/context)のみからマスク部を予測する事前学習タスクを解くことになる。 | ||
- 双方向ネットワーク | ||
MLM により双方向に依存するモデルを採用可能になったことから、BERT ではネットワークとして双方向性の Transformerアーキテクチャを採用した。すなわち self-attention による前後文脈取り込みと位置限局全結合による変換を繰り返すネットワークを用いている。 |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,24 @@ | ||
# Note | ||
``` | ||
guid: HFma{:j~;: | ||
notetype: AI-Vocabulary-Style | ||
``` | ||
|
||
### Tags | ||
``` | ||
``` | ||
|
||
## Front | ||
EMアルゴリズム | ||
|
||
## Back-EN | ||
Expectation–maximization algorithm | ||
|
||
## Back-FR | ||
Algorithme ésperance-maximisation | ||
|
||
## Reading | ||
EMアルゴリズム | ||
|
||
## Sentence | ||
EMアルゴリズムは反復法の一種であり、期待値(expectation, E) ステップと最大化 (maximization, M)ステップを交互に繰り替えすことで計算が進行する。Eステップでは、現在推定されている潜在変数の分布に基づいて、モデルの尤度の期待値を計算する。Mステップでは、E ステップで求まった尤度の期待値を最大化するようなパラメータを求める。M ステップで求まったパラメータは、次の E ステップで使われる潜在変数の分布を決定するために用いられる。 |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,24 @@ | ||
# Note | ||
``` | ||
guid: ElOpzZ*_o} | ||
notetype: AI-Vocabulary-Style | ||
``` | ||
|
||
### Tags | ||
``` | ||
``` | ||
|
||
## Front | ||
F値 / F1 | ||
|
||
## Back-EN | ||
F-score / F1-score (Harmonic mean of Precision and Recall) | ||
|
||
## Back-FR | ||
F-score / f1-score (moyenne harmonique de précision et rappel | ||
|
||
## Reading | ||
Fち / F1 | ||
|
||
## Sentence | ||
F値 = 2 * (敏感度 recall * 適合度 precision) / (敏感度 recall + 適合度 precision) |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,24 @@ | ||
# Note | ||
``` | ||
guid: rTI>,3SU1^ | ||
notetype: AI-Vocabulary-Style | ||
``` | ||
|
||
### Tags | ||
``` | ||
``` | ||
|
||
## Front | ||
F分布 | ||
|
||
## Back-EN | ||
F-distribution | ||
|
||
## Back-FR | ||
Loi de Fisher | ||
|
||
## Reading | ||
エフぶんぷ | ||
|
||
## Sentence | ||
|
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,24 @@ | ||
# Note | ||
``` | ||
guid: HlX;3T+Sre | ||
notetype: AI-Vocabulary-Style | ||
``` | ||
|
||
### Tags | ||
``` | ||
``` | ||
|
||
## Front | ||
GELU関数 / ガウス誤差線形ユニット | ||
|
||
## Back-EN | ||
GELU (Gaussian Error Linear Unit) function | ||
|
||
## Back-FR | ||
Redresseur GeLU (fonction Unité Linéaire Gaussian) | ||
|
||
## Reading | ||
GELUかんすう / ガウスごさせんけいユニット | ||
|
||
## Sentence | ||
|
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,24 @@ | ||
# Note | ||
``` | ||
guid: gkXR/]Q/Pw | ||
notetype: AI-Vocabulary-Style | ||
``` | ||
|
||
### Tags | ||
``` | ||
``` | ||
|
||
## Front | ||
GPT-3 モデル | ||
|
||
## Back-EN | ||
GPT-3 model (Generative Pretrained Transformer) | ||
|
||
## Back-FR | ||
Modèle GPT-3 (transformer génératif pré-entraîné) | ||
|
||
## Reading | ||
GPT-3 モデル | ||
|
||
## Sentence | ||
GPT-2の後継モデルである GPT-3 は、教師なしの Transformer 言語モデルである。 GPT-3 は 2020 年 5 月に初めて紹介された。 OpenAI によると、GPT-3 には 1,750 億個のパラメータが含まれ、GPT-2(パラメータ数 15 億個)より 2 桁大きい。 |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,24 @@ | ||
# Note | ||
``` | ||
guid: nlEC,})D;= | ||
notetype: AI-Vocabulary-Style | ||
``` | ||
|
||
### Tags | ||
``` | ||
``` | ||
|
||
## Front | ||
K平均法 | ||
|
||
## Back-EN | ||
K-means clustering | ||
|
||
## Back-FR | ||
Clustering k-means | ||
|
||
## Reading | ||
Kへいきんほう | ||
|
||
## Sentence | ||
k-means法は、まずデータを適当なクラスタに分けた後、クラスタの平均を用いてうまい具合にデータがわかれるように調整させていくアルゴリズムです。 |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,24 @@ | ||
# Note | ||
``` | ||
guid: Di#Ybbh9l; | ||
notetype: AI-Vocabulary-Style | ||
``` | ||
|
||
### Tags | ||
``` | ||
``` | ||
|
||
## Front | ||
K近傍法 | ||
|
||
## Back-EN | ||
K-Nearest Neighbors (k-NN) | ||
|
||
## Back-FR | ||
K-neaarest voisins (K-NN) | ||
|
||
## Reading | ||
ケイきんぼうほう(k-NN) | ||
|
||
## Sentence | ||
k近傍法は、特徴空間における最も近い訓練例に基づいた分類の手法であり、パターン認識でよく使われる。 |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,24 @@ | ||
# Note | ||
``` | ||
guid: dfT_c*[L^o | ||
notetype: AI-Vocabulary-Style | ||
``` | ||
|
||
### Tags | ||
``` | ||
``` | ||
|
||
## Front | ||
L1正則化 | ||
|
||
## Back-EN | ||
L1 regularization (LASSO, least absolute shrinkage and selection operator) | ||
|
||
## Back-FR | ||
L1 régularisation | ||
|
||
## Reading | ||
L1 せいそくか | ||
|
||
## Sentence | ||
ラッソ回帰(ラッソかいき、least absolute shrinkage and selection operator、Lasso、LASSO)は、変数選択と正則化の両方を実行し、生成する統計モデルの予測精度と解釈可能性を向上させる回帰分析手法。 |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,24 @@ | ||
# Note | ||
``` | ||
guid: ui~TzQSY:! | ||
notetype: AI-Vocabulary-Style | ||
``` | ||
|
||
### Tags | ||
``` | ||
``` | ||
|
||
## Front | ||
L2正則化 | ||
|
||
## Back-EN | ||
L2 regularization (Tikonov, Ridge regression) | ||
|
||
## Back-FR | ||
L2 régularisation (Tikonov, régression de la crête) | ||
|
||
## Reading | ||
L2 せいそくか | ||
|
||
## Sentence | ||
リッジ回帰(リッジかいき、Ridge regression)は、独立変数が強く相関している場合に、重回帰モデルの係数を推定する方法。計量経済学、化学、工学などの分野で使用されている。 |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,24 @@ | ||
# Note | ||
``` | ||
guid: hoh>//(5s6 | ||
notetype: AI-Vocabulary-Style | ||
``` | ||
|
||
### Tags | ||
``` | ||
``` | ||
|
||
## Front | ||
Leaky ReLU関数 | ||
|
||
## Back-EN | ||
Leaky ReLU (Rectified Linear Unit) function | ||
|
||
## Back-FR | ||
Redresseur Leaky ReLU (fonction Unité Linéaire Rectifiée) | ||
|
||
## Reading | ||
Leaky ReLUかんすう | ||
|
||
## Sentence | ||
|
Oops, something went wrong.