<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Kaggle Note</title>
    <link>https://kagglenote.com/</link>
    <description>Recent content on Kaggle Note</description>
    <generator>Hugo -- gohugo.io</generator>
    <language>ja</language>
    <copyright>&amp;copy; forest-hill 2020</copyright>
    <lastBuildDate>Sun, 23 Jul 2023 18:23:23 +0900</lastBuildDate>
    
	<atom:link href="https://kagglenote.com/index.xml" rel="self" type="application/rss+xml" />
    
    
    <item>
      <title>OpenAPIから自動生成したモデルでError: Undefined name &#39;truefalse&#39;.</title>
      <link>https://kagglenote.com/misc/openapi-truefalse-error/</link>
      <pubDate>Sun, 23 Jul 2023 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/openapi-truefalse-error/</guid>
      <description>問題 Flutterで開発している時にopenapi-generatorを使用してOpenAPI Schemaからモデルファイルを生成してビルド</description>
    </item>
    
    <item>
      <title>multiprocessingを使ったファイル読み込みの並列化</title>
      <link>https://kagglenote.com/misc/parallel_read_file/</link>
      <pubDate>Wed, 26 Apr 2023 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/parallel_read_file/</guid>
      <description>はじめに 今回はmultiprocessingを用いたファイル読み込みの並列化についてです。 直列で読み込むのに時間がかかるときに使えます。 ダミ</description>
    </item>
    
    <item>
      <title>scipyを用いたサーフェースフィッティング</title>
      <link>https://kagglenote.com/ml-tips/surface_fitting/</link>
      <pubDate>Fri, 03 Mar 2023 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/surface_fitting/</guid>
      <description>はじめに 今回はscipyを用いたサーフェスフィッティング（曲面近似）を行なっていきます。基本的にはカーブフィッティングと同じことをやればOK</description>
    </item>
    
    <item>
      <title>scipyのcurve_fitを使った関数近似</title>
      <link>https://kagglenote.com/ml-tips/curve_fitting/</link>
      <pubDate>Sat, 25 Feb 2023 20:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/curve_fitting/</guid>
      <description>はじめに 今回はscipy.optimize.curve_fitを用いて色々な関数のカーブフィッティングをしていきます。 ライブラリ import numpy as np import matplotlib.pyplot</description>
    </item>
    
    <item>
      <title>【matplotlib】箱ひげ図に折れ線グラフを追加</title>
      <link>https://kagglenote.com/misc/matplotlib_box_line/</link>
      <pubDate>Sat, 25 Feb 2023 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/matplotlib_box_line/</guid>
      <description>はじめに 今回は箱ひげ図に中央値の折れ線グラフを追加する方法です。 実装 ライブラリ import numpy as np import pandas as pd import matplotlib.pyplot as plt ダミーデータ df = [] sigma = 1 for i in range(5): df_tmp = pd.DataFrame({</description>
    </item>
    
    <item>
      <title>pandasのgroupbyにおける型別速度調査</title>
      <link>https://kagglenote.com/misc/pandas-groupby-speed/</link>
      <pubDate>Tue, 24 May 2022 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/pandas-groupby-speed/</guid>
      <description>はじめに pandasで集計をする際に必須なgroupby。 分析をするときには気にならないことが多いですが、システムとして運用する時は実行速度</description>
    </item>
    
    <item>
      <title>【python】bottleneckの速度調査</title>
      <link>https://kagglenote.com/misc/bottleneck/</link>
      <pubDate>Fri, 20 May 2022 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/bottleneck/</guid>
      <description>はじめに Fast NumPy array functions written in C. ドキュメントに記述されている通り、bottleneckはNumpyの一部の関数をC言語で書いたものであり、Cで書かれて</description>
    </item>
    
    <item>
      <title>sklearnのIsolationForestを用いた異常検知</title>
      <link>https://kagglenote.com/ml-tips/isolationforest/</link>
      <pubDate>Sun, 01 May 2022 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/isolationforest/</guid>
      <description>Isolation Forestとは Isolation Forestとは有名な異常検知手法の一つです。 Forestという名前からもわかるように決定木の仕組みを使って異常検知を行</description>
    </item>
    
    <item>
      <title>Darknetで学習をするまでの手順</title>
      <link>https://kagglenote.com/ml-tips/yolov4-training/</link>
      <pubDate>Sat, 01 Jan 2022 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/yolov4-training/</guid>
      <description>概要 今回は物体検出でよくお世話になるDarknetを用いてYolov4を学習するまでの手順を説明していきます。環境はUbuntuを想定してい</description>
    </item>
    
    <item>
      <title>OpenCVのonMouseを使った座標取得プログラム</title>
      <link>https://kagglenote.com/misc/get-point-with-opencv/</link>
      <pubDate>Sat, 16 Oct 2021 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/get-point-with-opencv/</guid>
      <description>概要 画像処理関連のことをしていると画像のここの座標を取得したい！！！という場面が結構出てきますよね。 今回は視覚的に矩形の座標を取得したいと思</description>
    </item>
    
    <item>
      <title>レコメンデーションまとめ書き</title>
      <link>https://kagglenote.com/ml-tips/recommend_memo/</link>
      <pubDate>Fri, 16 Jul 2021 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/recommend_memo/</guid>
      <description>概要 レコメンデーションとはECサイトなどでおすすめしてくる「あれ」である。 レコメンデーションの手法について調べたのでここにざっとまとめる。 手</description>
    </item>
    
    <item>
      <title>時系列分析における記述統計に関するメモ的まとめ</title>
      <link>https://kagglenote.com/ml-tips/ts_memo/</link>
      <pubDate>Sat, 10 Jul 2021 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/ts_memo/</guid>
      <description>概要 今回は時系列分析における最初のステップである記述統計について述べていきます。 述べるというよりは、ひとまずメモ的にまとめて、個々の手法につ</description>
    </item>
    
    <item>
      <title>pythonを使ったX-means法</title>
      <link>https://kagglenote.com/ml-tips/xmeans/</link>
      <pubDate>Thu, 03 Jun 2021 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/xmeans/</guid>
      <description>概要 この記事ではpythonを使ってX-means法を実装していきます。ライブラリはpyclusteringを用いています。 X-means法</description>
    </item>
    
    <item>
      <title>MNISTで1だけ取り出す</title>
      <link>https://kagglenote.com/ml-tips/mnist_only_1/</link>
      <pubDate>Wed, 19 May 2021 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/mnist_only_1/</guid>
      <description>概要 MNISTは言わずと知れた文字画像セットですね。それを使う上で「1」だけ喉から手が出るほどほしい！！！となることがあるかと思います(大袈</description>
    </item>
    
    <item>
      <title>KerasによるVAEの実装</title>
      <link>https://kagglenote.com/ml-tips/keras-vae/</link>
      <pubDate>Tue, 18 May 2021 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/keras-vae/</guid>
      <description>概要 Kerasを使ってVAE(Variational Autoencoder)の実装を行なっていきます。 この記事はこのチュートリアルをベースに</description>
    </item>
    
    <item>
      <title>CSSフレームワーク使ってるのにスマホの表示が小さい</title>
      <link>https://kagglenote.com/misc/cssframework_for_smartphone/</link>
      <pubDate>Thu, 06 May 2021 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/cssframework_for_smartphone/</guid>
      <description>問題点 bootstrapやbulmaなどが挙げられるCSSフレームワーク便利ですよね。しかし先日bulmaを使っているときに、スマホでの表示</description>
    </item>
    
    <item>
      <title>Bulma ExtensionsをCDNから使う</title>
      <link>https://kagglenote.com/misc/bulma_extensions_cdn/</link>
      <pubDate>Thu, 29 Apr 2021 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/bulma_extensions_cdn/</guid>
      <description>概要 Bulma ExtensionsをCDNから使う。タイトル通りである。特にhtmlファイルへの記述方法がメイン。 方法 以下のような感じである。 &amp;lt;link rel=&amp;#34;stylesheet&amp;#34; href=&amp;#34;https://cdn.jsdelivr.net/npm/bulma-extensions@4.0.0/dist/css/bulma-extensions.min.css&amp;#34;&amp;gt;</description>
    </item>
    
    <item>
      <title>【zsh: abort labelme】 Macでlabelmeを立ち上げようとしたらエラーがでる件</title>
      <link>https://kagglenote.com/misc/labelme_abort/</link>
      <pubDate>Wed, 14 Apr 2021 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/labelme_abort/</guid>
      <description>エラー内容 Macでlabelmeを立ち上げようとしたところ、以下のようなエラーが出た。 ***** 長いので省略 ***** zsh: abort labelme 解決方法 筆者の場合、これで解決で</description>
    </item>
    
    <item>
      <title>【python】multiprocessingでのクラス内変数の共有</title>
      <link>https://kagglenote.com/misc/multiprocessing_share_value_in_class/</link>
      <pubDate>Mon, 05 Apr 2021 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/multiprocessing_share_value_in_class/</guid>
      <description>はじめに pythonのおける並列処理の有名なライブラリにmultiprocessingというものがあります。このmultiprocessin</description>
    </item>
    
    <item>
      <title>【python】A/Bテストにおける統計検定-離散値編-</title>
      <link>https://kagglenote.com/misc/abtest_discrete/</link>
      <pubDate>Thu, 18 Mar 2021 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/abtest_discrete/</guid>
      <description>はじめに A/BテストといえばWebマーケティング戦略で最も使われる手法の一つです。一方でA/Bテストの際にしっかりと統計的な考え方を用いない</description>
    </item>
    
    <item>
      <title>【scikit-learn】 2値分類の際のpredictで任意の閾値を設定する</title>
      <link>https://kagglenote.com/ml-tips/sklearn-custom-threshold/</link>
      <pubDate>Fri, 12 Mar 2021 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/sklearn-custom-threshold/</guid>
      <description>はじめに scikit-learnにおける2値分類でpredictをするとデフォルトの閾値0.5で分類されますよね。今回はこの閾値を任意で設定</description>
    </item>
    
    <item>
      <title>コーエンのdをpythonで求める。</title>
      <link>https://kagglenote.com/misc/cohens-d/</link>
      <pubDate>Wed, 10 Feb 2021 12:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/cohens-d/</guid>
      <description>はじめに 今回は効果量で用いられるコーエンのd（Cohen’s d）をpythonで求めていきます。 コーエンのdとは？ コーエンのdとは2つのグル</description>
    </item>
    
    <item>
      <title>pythonでハミング距離を計算する</title>
      <link>https://kagglenote.com/misc/hamming-distance/</link>
      <pubDate>Tue, 09 Feb 2021 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/hamming-distance/</guid>
      <description>はじめに ここではハミング距離とはなにかの説明と、pythonでの計算方法を紹介します。 ハミング距離（Hamming Distance） ハミング</description>
    </item>
    
    <item>
      <title>pythonによるコサイン類似度の計算</title>
      <link>https://kagglenote.com/misc/cosine-similarity/</link>
      <pubDate>Sun, 07 Feb 2021 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/cosine-similarity/</guid>
      <description>はじめに ここではnumpyを利用してコサイン類似度の計算を行います。 コサイン類似度（Cosine Similarity）とは コサイン類似度とは</description>
    </item>
    
    <item>
      <title>【python】ユークリッド距離, マンハッタン距離, チェビシェフ距離, ミンコフスキー距離</title>
      <link>https://kagglenote.com/misc/various-distance/</link>
      <pubDate>Wed, 03 Feb 2021 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/various-distance/</guid>
      <description>はじめに 距離というと2点間の距離を真っ先に思いつきますが、世の中にはさまざまな距離の定義が存在します。 ここではユークリッド距離、マンハッタン</description>
    </item>
    
    <item>
      <title>jupyter-notebookでwarningsを非表示にする</title>
      <link>https://kagglenote.com/misc/notebook-warnings/</link>
      <pubDate>Mon, 21 Dec 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/notebook-warnings/</guid>
      <description>概要 jupyternote-bookを使ってるときに出てくるwarningsがうざったいので、非表示にする。 方法 import warnings warnings.filterwarnings(&amp;#39;ignore&amp;#39;) 以上。</description>
    </item>
    
    <item>
      <title>pandasで他のカラムを条件に、新しいカラムを追加する</title>
      <link>https://kagglenote.com/misc/pandas_create_new_column_based_on_other_column/</link>
      <pubDate>Wed, 09 Dec 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/pandas_create_new_column_based_on_other_column/</guid>
      <description>概要 pandasでデータの前処理をしている時、現在あるカラムを条件に使って新しいカラムを作成したい時ってありますよね。今回はその方法を紹介し</description>
    </item>
    
    <item>
      <title>EfficientNetを用いた画像分類モデルを学習させるまで。</title>
      <link>https://kagglenote.com/ml-tips/efficientnet/</link>
      <pubDate>Sat, 05 Dec 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/efficientnet/</guid>
      <description>概要 EfficientNetを用いた画像分類を行っていきます。この記事で実際に紹介するものは以下の通りです。 EfficientNetのインス</description>
    </item>
    
    <item>
      <title>numpyで最頻値を求める</title>
      <link>https://kagglenote.com/misc/numpy-mode/</link>
      <pubDate>Fri, 04 Dec 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/numpy-mode/</guid>
      <description>概要 numpyで最頻値を求める方法です。 方法 np.unique()を使います。 import numpy as np arr = np.array([0, 1, 2, 3, 4, 1, 2, 2, 2, 2, 2, 3, 4]) unique, freq = np.unique(arr, return_counts=True) #return</description>
    </item>
    
    <item>
      <title>PandasのNaNをif文で判定する</title>
      <link>https://kagglenote.com/misc/pandas_nan_judge/</link>
      <pubDate>Thu, 05 Nov 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/pandas_nan_judge/</guid>
      <description>概要 Pandasを使うときに出てくるNaNをif文で判定する方法を紹介します。 準備 以下のcsvファイルがあると想定します。ここではtest.</description>
    </item>
    
    <item>
      <title>/var/lib/dpkg/lock-frontend が取得できませんでした</title>
      <link>https://kagglenote.com/misc/lock-frontend/</link>
      <pubDate>Fri, 23 Oct 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/lock-frontend/</guid>
      <description>問題 以下のようなエラーが出てしまう E: ロック /var/lib/dpkg/lock-frontend が取得できませんでした - open (11: リソースが一時的に利用できません) 解決策 以下のコマンドを打つと僕の場</description>
    </item>
    
    <item>
      <title>xpression camera &#43; zoomで遊んでみる</title>
      <link>https://kagglenote.com/misc/xpression-camera/</link>
      <pubDate>Wed, 07 Oct 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/xpression-camera/</guid>
      <description>はじめに COVID-19が流行して働き方が変わり、リモートワークがニューノーマルと考えられるようになりましたね。大学などでもzoomのような</description>
    </item>
    
    <item>
      <title>Pythonを使って0.1ごとにfor文を回す</title>
      <link>https://kagglenote.com/misc/python-for-decimal/</link>
      <pubDate>Fri, 02 Oct 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/python-for-decimal/</guid>
      <description>はじめに for文を使って小数点を繰り返したい時ってたまにありますよね。今回はそれを実現する方法を書いていきます。0~1を0.1刻みで繰り返す</description>
    </item>
    
    <item>
      <title>MacにPostgreSQLをインストールする</title>
      <link>https://kagglenote.com/misc/mac-postgresql-install/</link>
      <pubDate>Wed, 30 Sep 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/misc/mac-postgresql-install/</guid>
      <description>はじめに SQLの勉強のためにPostgreSQLを使いたいと思います。そこで自分のMacにPostgreSQLをインストールしていきたいと思</description>
    </item>
    
    <item>
      <title>OpenAI Gymの環境を自作して学習してみる</title>
      <link>https://kagglenote.com/ml-tips/my-environment-with-gym/</link>
      <pubDate>Sun, 20 Sep 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/my-environment-with-gym/</guid>
      <description>概要 自作方法 とりあえずこんな感じで書いていけばOK import gym class MyEnv(gym.Env): def __init__(self): ACTION_NUM=3 #アクションの数が3つの場合 self.action_space = gym.spaces.Discrete(ACTION_NUM) #状態が3つの時で上限と下限の設定と仮定 LOW=[0,0,0]</description>
    </item>
    
    <item>
      <title>【python】 Numpyで2点間の距離と角度を求める</title>
      <link>https://kagglenote.com/ml-tips/between-two-points/</link>
      <pubDate>Fri, 18 Sep 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/between-two-points/</guid>
      <description>概要 numpyを使って以下のような2点間の距離と角度を求めていきます。 距離 距離の計算ではnp.linalg.normを使います。 &amp;gt;&amp;gt;&amp;gt; import numpy as np &amp;gt;&amp;gt;&amp;gt;</description>
    </item>
    
    <item>
      <title>【Keras】Callback関数を自作する</title>
      <link>https://kagglenote.com/ml-tips/keras-callback/</link>
      <pubDate>Tue, 15 Sep 2020 12:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/keras-callback/</guid>
      <description>概要 KerasのCallback関数を自作します。 使い方 from keras.callbacks import Callback class MyCallback(Callback): def __init__(self): pass def on_epoch_begin(self, epoch, logs=None): pass def on_epoch_end(self, epoch, logs=None): pass def on_batch_begin(self, batch, logs=None): pass def on_batch_end(self, batch, logs=None): pass def on_train_begin(self, logs=None): pass def on_train_end(self, logs=None): pass 関数 処理タ</description>
    </item>
    
    <item>
      <title>【Keras】モデルの可視化</title>
      <link>https://kagglenote.com/ml-tips/keras-plt-model/</link>
      <pubDate>Sun, 13 Sep 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/keras-plt-model/</guid>
      <description>概要 Kerasで構築したニューラルネットワークのモデルを可視化するためにplot_modelを使ってみたいと思います。 keras.utils.plot_model(model, show_shapes=True) 全結合 import keras inputs = keras.layers.Input(shape=(64,)) x</description>
    </item>
    
    <item>
      <title>numpyで特定の値を削除する</title>
      <link>https://kagglenote.com/ml-tips/numpy-delete-specific-value/</link>
      <pubDate>Fri, 11 Sep 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/numpy-delete-specific-value/</guid>
      <description>やること 特定の数値の削除 条件に該当する数値の削除 array([-3, -2, -1, 0, 1, 2, 3]) # ↓ 0を削除する array([-3, -2, -1, 1, 2, 3]) array([-3, -2, -1, 0, 1, 2, 3]) # ↓ 負の数を削除する array([0, 1, 2, 3]) 特定の数</description>
    </item>
    
    <item>
      <title>【Keras】 CNNを用いたMNISTクラス分類</title>
      <link>https://kagglenote.com/ml-tips/keras-cnn/</link>
      <pubDate>Wed, 09 Sep 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/keras-cnn/</guid>
      <description>概要 KerasでCNNを構築してMNISTクラス分類を行っていきます ライブラリインポート import os import keras from keras.models import Sequential from keras.layers.convolutional import Conv2D, MaxPooling2D from keras.layers.core import Activation, Flatten, Dropout, Dense from keras.datasets import mnist from keras.optimizers</description>
    </item>
    
    <item>
      <title>疑似ラベリングの実装</title>
      <link>https://kagglenote.com/kaggle/pseudo-labeling-impl/</link>
      <pubDate>Mon, 07 Sep 2020 15:00:00 +0900</pubDate>
      
      <guid>https://kagglenote.com/kaggle/pseudo-labeling-impl/</guid>
      <description>疑似ラベリングとは学習時にテストデータの一部を予測したものを含めて学習し、このモデルを使って再度テストデータの全体を学習することです。以下に</description>
    </item>
    
    <item>
      <title>ブレンディングの重み決定手法</title>
      <link>https://kagglenote.com/kaggle/blending-based-on-oof/</link>
      <pubDate>Sat, 05 Sep 2020 15:00:00 +0900</pubDate>
      
      <guid>https://kagglenote.com/kaggle/blending-based-on-oof/</guid>
      <description>ブレンディングとは複数のモデルで予測した結果を合わせることでより良い結果を作り出すことです。 (モデル1の結果) * 0.3 + (モデル2の結果) * 0.4 +</description>
    </item>
    
    <item>
      <title>【DDQN】stable-baselinesで始める強化学習</title>
      <link>https://kagglenote.com/ml-tips/stable-baselines-ddqn/</link>
      <pubDate>Thu, 03 Sep 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/stable-baselines-ddqn/</guid>
      <description>やること stable-baselinesを使ってDDQNを動かします。環境としてはGymのCartPoleを利用していきます。 stable-</description>
    </item>
    
    <item>
      <title>【準備】stable-baselinesで始める強化学習</title>
      <link>https://kagglenote.com/ml-tips/stable-baselines-install/</link>
      <pubDate>Tue, 01 Sep 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/stable-baselines-install/</guid>
      <description>stable-baselinesとは stable-baselinesとは様々な強化学習アルゴリズムが実装されていて、誰でも気軽に強化学習を試</description>
    </item>
    
    <item>
      <title>【tensorflow】 YOLOv4をとりあえず動かしてみた</title>
      <link>https://kagglenote.com/ml-tips/yolov4/</link>
      <pubDate>Mon, 31 Aug 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/yolov4/</guid>
      <description>概要 物体検出のYOLOv4をとりあえず動かしてみる。 やってみる 実行環境 $ lsb_release -a No LSB modules are available. Distributor ID: Pop Description: Pop!_OS 20.04 LTS Release: 20.04 Codename: focal $ sudo lshw -class processor *-cpu description: CPU product: Intel(R) Core(TM) i7-8565U CPU @ 1.80GHz vendor:</description>
    </item>
    
    <item>
      <title>【Kears】 LSTM,GRUをとりあえず動かしてみる</title>
      <link>https://kagglenote.com/ml-tips/timeseries-nn-sin/</link>
      <pubDate>Sat, 29 Aug 2020 11:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/timeseries-nn-sin/</guid>
      <description>今回やること 時系列系のニューラルネットワークといえば、LSTM・GRUがぱっと思いつきます。今回はこれらをぱっと動かしてみたいと思います。 デ</description>
    </item>
    
    <item>
      <title>simpletransformersを使ったテキスト分類</title>
      <link>https://kagglenote.com/kaggle/simpletransformers-claassification/</link>
      <pubDate>Thu, 27 Aug 2020 15:00:00 +0900</pubDate>
      
      <guid>https://kagglenote.com/kaggle/simpletransformers-claassification/</guid>
      <description>自然言語処理の世界では2019年位から事前学習による学習モデルの使用が活発になっています。これはAttentionベースであるGoogleの</description>
    </item>
    
    <item>
      <title>YOLOv5動かしてみた</title>
      <link>https://kagglenote.com/ml-tips/yolov5/</link>
      <pubDate>Tue, 25 Aug 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/yolov5/</guid>
      <description>概要 物体検出でよく用いられるYOLOですが、その最新版（2020/08現在）であるYOLOv5を今回は動かしてみます。 やってみる 検証環境 試し</description>
    </item>
    
    <item>
      <title>【python】 bert-extractive-summarizerを使った文章要約</title>
      <link>https://kagglenote.com/ml-tips/bert-extractive-summarizer/</link>
      <pubDate>Sun, 23 Aug 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/bert-extractive-summarizer/</guid>
      <description>概要 今回は文章要約でとてもpythonの有用なライブラリbert-extractive-summarizerの使い方を簡単に紹介したいと思い</description>
    </item>
    
    <item>
      <title>fastaiを用いたセマンティックセグメンテーション</title>
      <link>https://kagglenote.com/kaggle/semantic-segmentation-with-fastai/</link>
      <pubDate>Fri, 21 Aug 2020 12:00:00 +0900</pubDate>
      
      <guid>https://kagglenote.com/kaggle/semantic-segmentation-with-fastai/</guid>
      <description>概要 今回紹介するnotebookはAerial_Drone_semantic_image_segmentation | Fastaiです。 このn</description>
    </item>
    
    <item>
      <title>【One-Hotエンコーディング】 scikit-learn vs pandas</title>
      <link>https://kagglenote.com/ml-tips/onehotencoder-sklearn-vs-pandas/</link>
      <pubDate>Wed, 19 Aug 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/onehotencoder-sklearn-vs-pandas/</guid>
      <description>one-hotエンコーディングとは？ one-hotエンコーディングとはカテゴリデータに対する処理の一つです。 カテゴリデータは数値データではな</description>
    </item>
    
    <item>
      <title>スタッキングライブラリvecstackの使い方</title>
      <link>https://kagglenote.com/kaggle/stacking-by-vecstack/</link>
      <pubDate>Mon, 17 Aug 2020 15:00:00 +0900</pubDate>
      
      <guid>https://kagglenote.com/kaggle/stacking-by-vecstack/</guid>
      <description>この記事ではスタッキングについてvecstackの使い方をrisを例として説明します。 スタッキングとは スタッキングは異なる予測器を1層2層&amp;</description>
    </item>
    
    <item>
      <title>逆翻訳を使ったテキストデータ水増し</title>
      <link>https://kagglenote.com/kaggle/back-translation/</link>
      <pubDate>Sat, 15 Aug 2020 15:00:00 +0900</pubDate>
      
      <guid>https://kagglenote.com/kaggle/back-translation/</guid>
      <description>テキスト分類を行う際、教師データを増やすために水増しをすることは非常に重要です。 テキストデータ水増しの手法として、Data Augmentation in NLPという記</description>
    </item>
    
    <item>
      <title>numpyで学習結果の多数決</title>
      <link>https://kagglenote.com/ml-tips/majority-vote/</link>
      <pubDate>Thu, 13 Aug 2020 15:00:00 +0900</pubDate>
      
      <guid>https://kagglenote.com/ml-tips/majority-vote/</guid>
      <description>複数の学習結果を多数決したい時、以下のような操作が必要になります。 これを実現するには転置してmapで各行に対して最頻値を求めればOKです。 results</description>
    </item>
    
    <item>
      <title>【入門】pythonを用いた自然言語処理</title>
      <link>https://kagglenote.com/kaggle/nlp-for-beginners/</link>
      <pubDate>Tue, 11 Aug 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/kaggle/nlp-for-beginners/</guid>
      <description>概要 今回紹介するnotebookはNatural Language Processing (NLP) 🧾 for Beginnersです。 この記事は自然言語処理初学者向けであり、自然言語処理に関す</description>
    </item>
    
    <item>
      <title>TF-IDFを使ったテキストベクトル化</title>
      <link>https://kagglenote.com/kaggle/text-vectorize-by-tfidf/</link>
      <pubDate>Sun, 09 Aug 2020 00:00:00 +0900</pubDate>
      
      <guid>https://kagglenote.com/kaggle/text-vectorize-by-tfidf/</guid>
      <description>テキスト分類問題等で使うことが多いテキストベクトル化について説明します。 1. テキストベクトル化 テキストベクトル化とは主に自然言語処理で使われる</description>
    </item>
    
    <item>
      <title>このサイトについて</title>
      <link>https://kagglenote.com/about/</link>
      <pubDate>Fri, 07 Aug 2020 20:13:29 +0900</pubDate>
      
      <guid>https://kagglenote.com/about/</guid>
      <description>このサイトでは機械学習に取り組んでいる中で自分たちが勉強した内容をまとめた記事を公開しています。 書いてる人 asmsuechanとyokopo</description>
    </item>
    
    <item>
      <title>【入門】Keras LSTMを用いたテキスト分類</title>
      <link>https://kagglenote.com/kaggle/keras-lstm-textclassifier/</link>
      <pubDate>Fri, 07 Aug 2020 18:23:23 +0900</pubDate>
      
      <guid>https://kagglenote.com/kaggle/keras-lstm-textclassifier/</guid>
      <description>概要 今回紹介するnotebookは[For Beginners] Tackling Toxic Using Kerasです。このnotebookではKerasのLSTMを用いたテキスト分類について書</description>
    </item>
    
  </channel>
</rss>