-
-
T/O
-
- 問1.1
> SGDでは並列化ができないことは留意する
- 問1.3
> SGDは等高線が円になっている場合はうまくいくが、楕円など歪んでいる場合(当方的でない)はジグザグに更新される
p(x ; Θ)・・・Θというパラメータが与えられた下でのp(x)の評価値
p(x | Θ)・・・Θという条件が与えられた下でのp(x)の評価値
データxが与えられたときのシータの確率はp(Θ …
-
まず、いくつかのクラスを実装したあと、さらに広げていこう。
* https://www.fpcomplete.com/blog/2017/06/readert-design-pattern
* https://twitter.com/hexirp_prixeh/status/984979486755008512
* https://twitter.com/hexirp_prixeh/st…
-
+ 深さに寄る重み付けの値と,分割の閾値を自動的に決めたい
+ ユーザのWDがどの層に多く分布しているかにより,何層目で分割しておくか決めたい
+ 実は閾値固定で重み付けを考えるだけでよい.(小数が出現するが)
+ 分布から何層目で切りたいか→重みの決定
よく分からんくなった...
-
DMからサポートが受けれるようにしたい
-
手打ちしたくないので、Pythonスクリプトで変換する
jaconv
-
-
OSCのmessageの構成検討をする
-
https://scrapbox.io/nikkie-memos/%22Why_Should_I_Trust_You%3F%22:_Explaining_the_Predictions_of_Any_Classifier
## まとめ
どのような分類器の予測も説明できる手法LIME (Local Interpretable Model-agnostic Explanations) を提…