3_2三种特征选择方法的对比.ipynb · likunzm/sklearn-ml - Gitee.com

4月12日模力方舟 AI 应用沙龙 · 杭州站报名开放，产研前线第一手干货，AI 开发者必冲！

forked from duanchenyang/sklearn-ml

加入 Gitee

与超过 1200万开发者一起发现、参与优秀开源项目，私有仓库也完全免费：）

文件

该仓库未声明开源许可证文件（LICENSE），使用请关注具体项目描述及其代码上游依赖。

克隆/下载

3_2三种特征选择方法的对比.ipynb 30.35 KB

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"name":"3_2三种特征选择方法的对比.ipynb","version":"0.3.2","provenance":[],"collapsed_sections":[]},"kernelspec":{"name":"python3","display_name":"Python 3"},"accelerator":"TPU"},"cells":[{"metadata":{"id":"PbPreUjCnBqA","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":35},"outputId":"31c2ed67-af77-4d08-a829-b4914693b689","executionInfo":{"status":"ok","timestamp":1546131898450,"user_tz":-480,"elapsed":4935,"user":{"displayName":"Sen Yang","photoUrl":"","userId":"00832503676208839570"}}},"cell_type":"code","source":["import pandas as pd\n","import numpy as np\n","data = pd.read_csv(r\"./digit recognizor.csv\")\n","X = data.iloc[:,1:]\n","y = data.iloc[:,0]\n","\n","print(X.shape)"],"execution_count":1,"outputs":[{"output_type":"stream","text":["(42000, 784)\n"],"name":"stdout"}]},{"metadata":{"id":"JDvgRRu9ngx5","colab_type":"text"},"cell_type":"markdown","source":["确定嵌入法选择特征的个数"]},{"metadata":{"id":"hf0eAnl7nU1y","colab_type":"code","colab":{}},"cell_type":"code","source":["from sklearn.feature_selection import SelectFromModel\n","from sklearn.ensemble import RandomForestClassifier as RFC\n","RFC_ = RFC(n_estimators =10,random_state=0)\n","X_embedded = SelectFromModel(RFC_,threshold=0.0045).fit_transform(X,y)\n"],"execution_count":0,"outputs":[]},{"metadata":{"id":"jD6BqL0Qnfxx","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":35},"outputId":"180a3280-3402-4d0c-f8f2-a9fd5d636b99","executionInfo":{"status":"ok","timestamp":1546131924132,"user_tz":-480,"elapsed":658,"user":{"displayName":"Sen Yang","photoUrl":"","userId":"00832503676208839570"}}},"cell_type":"code","source":["X_embedded.shape"],"execution_count":3,"outputs":[{"output_type":"execute_result","data":{"text/plain":["(42000, 61)"]},"metadata":{"tags":[]},"execution_count":3}]},{"metadata":{"id":"qncbQ8DwnouB","colab_type":"text"},"cell_type":"markdown","source":["过滤法选择特征"]},{"metadata":{"id":"p8kBEajDnvJx","colab_type":"code","colab":{}},"cell_type":"code","source":["xx = X.var().values\n"],"execution_count":0,"outputs":[]},{"metadata":{"id":"xY7r9mZHpfNh","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":3403},"outputId":"32ad6915-203f-4b7d-bd23-5c621b29395c","executionInfo":{"status":"ok","timestamp":1546132056270,"user_tz":-480,"elapsed":443,"user":{"displayName":"Sen Yang","photoUrl":"","userId":"00832503676208839570"}}},"cell_type":"code","source":["xx"],"execution_count":9,"outputs":[{"output_type":"execute_result","data":{"text/plain":["array([0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,\n","       0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,\n","       0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,\n","       3.22760590e-01, 2.64689018e+00, 1.11085714e+00, 1.92857143e-03,\n","       0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,\n","       0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,\n","       0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,\n","       0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,\n","       6.09523810e-03, 5.41186213e-02, 1.28065733e+00, 5.33793150e+00,\n","       9.74592647e+00, 1.06219147e+01, 2.49289953e+01, 3.37671294e+01,\n","       3.59425660e+01, 3.49309575e+01, 3.62943342e+01, 3.27911079e+01,\n","       3.04237672e+01, 2.95547817e+01, 2.00568398e+01, 1.19545659e+01,\n","       8.60419724e+00, 2.55465685e+00, 1.60530842e+00, 9.29377602e-01,\n","       0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,\n","       0.00000000e+00, 0.00000000e+00, 9.75238095e-02, 2.00238095e-02,\n","       8.38442376e-01, 6.23481035e-01, 9.48446647e+00, 2.93530906e+01,\n","       5.67872968e+01, 1.00230634e+02, 1.65873698e+02, 2.52738939e+02,\n","       3.76018084e+02, 5.14426094e+02, 6.50102357e+02, 7.26608631e+02,\n","       7.20549617e+02, 6.49594181e+02, 5.18814455e+02, 3.94092376e+02,\n","       2.35038093e+02, 1.14843923e+02, 5.61521180e+01, 1.70331842e+01,\n","       6.48083740e+00, 1.08821717e+00, 0.00000000e+00, 0.00000000e+00,\n","       0.00000000e+00, 0.00000000e+00, 4.76236337e-01, 2.75489198e-01,\n","       1.29297595e+00, 1.26446724e+01, 4.28153306e+01, 9.43924926e+01,\n","       2.15866693e+02, 4.40079141e+02, 7.26679321e+02, 1.10776068e+03,\n","       1.49073923e+03, 1.93993499e+03, 2.31552090e+03, 2.59022407e+03,\n","       2.49637882e+03, 2.22627799e+03, 1.80800104e+03, 1.31417708e+03,\n","       7.95887557e+02, 4.41320506e+02, 1.99165391e+02, 7.95865424e+01,\n","       3.10264989e+01, 2.26837146e+00, 3.48595238e-01, 0.00000000e+00,\n","       0.00000000e+00, 3.43809524e-02, 1.15447633e-01, 1.25289059e+00,\n","       1.15530325e+01, 6.16749834e+01, 1.71059884e+02, 4.37266473e+02,\n","       9.14503323e+02, 1.61733281e+03, 2.58238559e+03, 3.73319320e+03,\n","       4.97101576e+03, 6.18285701e+03, 7.12176198e+03, 7.61902605e+03,\n","       7.35140128e+03, 6.59732933e+03, 5.42159265e+03, 4.14622275e+03,\n","       2.74210919e+03, 1.60716526e+03, 8.30092375e+02, 3.85524893e+02,\n","       1.44471061e+02, 2.98361218e+01, 5.45472846e+00, 0.00000000e+00,\n","       0.00000000e+00, 0.00000000e+00, 7.16499590e-01, 5.19710545e+00,\n","       6.12264995e+01, 2.53650903e+02, 6.51218510e+02, 1.32512948e+03,\n","       2.39625599e+03, 3.90432573e+03, 5.65248822e+03, 7.50750831e+03,\n","       9.28013155e+03, 1.07098605e+04, 1.15505424e+04, 1.18353399e+04,\n","       1.16728383e+04, 1.10553420e+04, 9.77051209e+03, 8.00603359e+03,\n","       5.90601149e+03, 3.83759619e+03, 2.22900745e+03, 1.15756773e+03,\n","       5.07963705e+02, 1.22007799e+02, 1.78286793e+01, 7.71428571e-03,\n","       0.00000000e+00, 3.80952381e-04, 2.37199928e+00, 3.95713081e+01,\n","       2.07427935e+02, 6.06480056e+02, 1.37944393e+03, 2.64652318e+03,\n","       4.39063213e+03, 6.55335113e+03, 8.64850053e+03, 1.03924767e+04,\n","       1.16602029e+04, 1.23720092e+04, 1.25944602e+04, 1.26223718e+04,\n","       1.26261428e+04, 1.25617677e+04, 1.20570623e+04, 1.07029270e+04,\n","       8.66359114e+03, 6.04366459e+03, 3.77715738e+03, 2.05026195e+03,\n","       9.47173606e+02, 3.04430342e+02, 5.49548377e+01, 3.50648111e+00,\n","       0.00000000e+00, 1.47595606e+00, 1.73365714e+01, 1.03708159e+02,\n","       4.49220215e+02, 1.11598051e+03, 2.31261700e+03, 4.05981199e+03,\n","       6.27112094e+03, 8.61258120e+03, 1.04088978e+04, 1.15863594e+04,\n","       1.20861191e+04, 1.22313676e+04, 1.22256615e+04, 1.21230025e+04,\n","       1.21304108e+04, 1.22345097e+04, 1.22810693e+04, 1.17587829e+04,\n","       1.01382911e+04, 7.63237251e+03, 4.91736987e+03, 2.72324144e+03,\n","       1.22087326e+03, 4.16952121e+02, 7.84619988e+01, 2.69471891e+00,\n","       6.68809524e-02, 1.14742637e+01, 7.06094700e+01, 2.58579123e+02,\n","       7.58812237e+02, 1.66236697e+03, 3.16372296e+03, 5.28061009e+03,\n","       7.80859675e+03, 1.00912727e+04, 1.16171690e+04, 1.22797788e+04,\n","       1.23833077e+04, 1.24581779e+04, 1.23866060e+04, 1.23021558e+04,\n","       1.22796752e+04, 1.23185424e+04, 1.24370140e+04, 1.21441547e+04,\n","       1.08800641e+04, 8.45510124e+03, 5.53590659e+03, 2.92299426e+03,\n","       1.27293575e+03, 4.39632397e+02, 7.77464211e+01, 2.35152706e+00,\n","       1.41387108e+00, 1.83873300e+01, 1.01219976e+02, 3.59094119e+02,\n","       9.16376867e+02, 1.90547923e+03, 3.64144404e+03, 6.06890402e+03,\n","       8.86696591e+03, 1.10299773e+04, 1.21997003e+04, 1.24677668e+04,\n","       1.23878202e+04, 1.20816759e+04, 1.18414911e+04, 1.18243379e+04,\n","       1.19196049e+04, 1.22058487e+04, 1.24040782e+04, 1.21013351e+04,\n","       1.08905421e+04, 8.53288188e+03, 5.50599673e+03, 2.77180472e+03,\n","       1.06085951e+03, 3.12689774e+02, 5.19949129e+01, 5.02228364e+00,\n","       3.12121001e+00, 1.92336286e+01, 1.00196288e+02, 3.33760276e+02,\n","       8.42723536e+02, 1.87713749e+03, 3.80614513e+03, 6.46513903e+03,\n","       9.41022768e+03, 1.14543460e+04, 1.22494028e+04, 1.22551059e+04,\n","       1.15972630e+04, 1.07756156e+04, 1.05723907e+04, 1.10540147e+04,\n","       1.14904415e+04, 1.20641837e+04, 1.22340857e+04, 1.19219747e+04,\n","       1.04672465e+04, 8.02282693e+03, 5.11996418e+03, 2.49340573e+03,\n","       8.29185839e+02, 2.02728469e+02, 3.65252982e+01, 4.73545802e+00,\n","       6.54689762e-01, 1.10762921e+01, 8.21810900e+01, 2.45993138e+02,\n","       6.88037918e+02, 1.77097595e+03, 3.89510214e+03, 6.82414020e+03,\n","       9.88524752e+03, 1.16866740e+04, 1.22052249e+04, 1.18503310e+04,\n","       1.08057141e+04, 1.00029368e+04, 1.05613005e+04, 1.13459087e+04,\n","       1.16813954e+04, 1.21153444e+04, 1.22013911e+04, 1.16257849e+04,\n","       9.77524122e+03, 7.24039730e+03, 4.72965127e+03, 2.39611574e+03,\n","       7.37971166e+02, 1.05994299e+02, 2.09059137e+01, 7.86044778e-01,\n","       6.32595238e-01, 6.32770971e+00, 4.96597411e+01, 1.73335928e+02,\n","       5.64355135e+02, 1.79246896e+03, 4.19852946e+03, 7.42330328e+03,\n","       1.03680620e+04, 1.19241275e+04, 1.22167302e+04, 1.17312041e+04,\n","       1.08331969e+04, 1.06956425e+04, 1.18765965e+04, 1.23125550e+04,\n","       1.21036991e+04, 1.23648277e+04, 1.23532159e+04, 1.12659827e+04,\n","       9.01309194e+03, 6.65304923e+03, 4.53230600e+03, 2.53236771e+03,\n","       8.13317532e+02, 7.57360821e+01, 1.22663918e+01, 9.49856083e-01,\n","       2.43809524e-02, 2.05312854e+00, 2.37737153e+01, 1.15295050e+02,\n","       5.23151198e+02, 2.00356402e+03, 4.71208413e+03, 8.02210576e+03,\n","       1.07142826e+04, 1.19844734e+04, 1.21443066e+04, 1.17352672e+04,\n","       1.13501890e+04, 1.18058310e+04, 1.29305253e+04, 1.24664835e+04,\n","       1.21090771e+04, 1.25723716e+04, 1.24760936e+04, 1.09595385e+04,\n","       8.66572810e+03, 6.56048630e+03, 4.59945505e+03, 2.74090145e+03,\n","       9.79975029e+02, 1.02467830e+02, 1.16955902e+01, 2.92035599e+00,\n","       0.00000000e+00, 8.53042930e-01, 8.80331299e+00, 7.52414185e+01,\n","       5.33884018e+02, 2.30333924e+03, 5.22384299e+03, 8.41596660e+03,\n","       1.08537671e+04, 1.19270178e+04, 1.19856397e+04, 1.17039418e+04,\n","       1.16715555e+04, 1.24445986e+04, 1.29618550e+04, 1.20230391e+04,\n","       1.20377425e+04, 1.26434659e+04, 1.24012680e+04, 1.08759560e+04,\n","       8.87523588e+03, 6.85569402e+03, 4.84378377e+03, 2.91796192e+03,\n","       1.10844668e+03, 1.39672375e+02, 8.86639995e+00, 7.94981734e-02,\n","       0.00000000e+00, 0.00000000e+00, 4.63950502e+00, 8.33393904e+01,\n","       6.29744755e+02, 2.69185746e+03, 5.62185375e+03, 8.53435292e+03,\n","       1.06354679e+04, 1.16397177e+04, 1.17410232e+04, 1.17248709e+04,\n","       1.20446633e+04, 1.27114555e+04, 1.27125078e+04, 1.20008852e+04,\n","       1.24098689e+04, 1.26745406e+04, 1.20741286e+04, 1.08898533e+04,\n","       9.21323231e+03, 7.22598679e+03, 5.04689745e+03, 2.90913735e+03,\n","       1.12277645e+03, 1.91078386e+02, 2.57469149e+01, 1.29179123e+00,\n","       0.00000000e+00, 7.69744852e-02, 1.03197870e+01, 1.11053766e+02,\n","       8.11564748e+02, 3.07386387e+03, 5.91065656e+03, 8.36008213e+03,\n","       1.01408287e+04, 1.10632136e+04, 1.12910607e+04, 1.14976813e+04,\n","       1.20735231e+04, 1.27502876e+04, 1.26987222e+04, 1.23901805e+04,\n","       1.26170244e+04, 1.23916950e+04, 1.17483338e+04, 1.09217037e+04,\n","       9.48760960e+03, 7.35171522e+03, 5.01095743e+03, 2.72992609e+03,\n","       1.04916705e+03, 2.25941630e+02, 3.05993271e+01, 2.87316363e+00,\n","       0.00000000e+00, 2.28039253e-01, 1.74352649e+01, 1.69998742e+02,\n","       1.10767541e+03, 3.43926495e+03, 6.05208363e+03, 8.08344097e+03,\n","       9.50906305e+03, 1.03073030e+04, 1.06715518e+04, 1.10973680e+04,\n","       1.17642567e+04, 1.22940460e+04, 1.25222837e+04, 1.24838281e+04,\n","       1.23538616e+04, 1.20985659e+04, 1.17177753e+04, 1.10570877e+04,\n","       9.47381642e+03, 7.19094489e+03, 4.63491474e+03, 2.42432309e+03,\n","       9.17534777e+02, 2.34119338e+02, 3.27687618e+01, 1.59861539e+00,\n","       1.47639388e-01, 4.47934602e-01, 2.76028871e+01, 2.71942488e+02,\n","       1.45371251e+03, 3.83133451e+03, 6.28631187e+03, 8.14612790e+03,\n","       9.30917203e+03, 1.00179199e+04, 1.05639229e+04, 1.09925571e+04,\n","       1.13187529e+04, 1.19116163e+04, 1.23480093e+04, 1.23440430e+04,\n","       1.21692394e+04, 1.21290528e+04, 1.19460348e+04, 1.11011974e+04,\n","       9.19667530e+03, 6.71976236e+03, 4.18683656e+03, 2.12415117e+03,\n","       8.18965786e+02, 2.22857925e+02, 2.54140760e+01, 1.36995713e+00,\n","       0.00000000e+00, 1.81578450e+00, 4.96345920e+01, 3.80903626e+02,\n","       1.68403918e+03, 4.10905800e+03, 6.66129020e+03, 8.56235623e+03,\n","       9.76878976e+03, 1.05714566e+04, 1.11195093e+04, 1.14535438e+04,\n","       1.17907581e+04, 1.21204054e+04, 1.24479360e+04, 1.23900112e+04,\n","       1.22489975e+04, 1.23606302e+04, 1.19785927e+04, 1.06491638e+04,\n","       8.49640625e+03, 5.95805507e+03, 3.53263357e+03, 1.70287256e+03,\n","       6.43673157e+02, 1.79133677e+02, 2.84028991e+01, 6.28175963e-01,\n","       0.00000000e+00, 1.23625493e+00, 5.65580767e+01, 4.36617659e+02,\n","       1.67611596e+03, 3.91825583e+03, 6.61848502e+03, 8.93197399e+03,\n","       1.05084494e+04, 1.14328933e+04, 1.19110401e+04, 1.21686631e+04,\n","       1.23576844e+04, 1.24269771e+04, 1.24145723e+04, 1.24011494e+04,\n","       1.23905027e+04, 1.22793811e+04, 1.14389559e+04, 9.61909799e+03,\n","       7.11272858e+03, 4.61189510e+03, 2.55069845e+03, 1.22448673e+03,\n","       4.85654757e+02, 1.30709778e+02, 1.54997227e+01, 5.95238095e-02,\n","       2.43809524e-02, 5.50986923e-01, 4.58334020e+01, 3.50464999e+02,\n","       1.27200240e+03, 3.19694301e+03, 5.84711073e+03, 8.57024623e+03,\n","       1.06775424e+04, 1.19000918e+04, 1.23555360e+04, 1.24402093e+04,\n","       1.23586522e+04, 1.21992581e+04, 1.22280936e+04, 1.23162507e+04,\n","       1.22510127e+04, 1.16314635e+04, 1.00849238e+04, 7.71018098e+03,\n","       5.16732671e+03, 2.99318809e+03, 1.58471510e+03, 7.70678750e+02,\n","       2.90603130e+02, 7.36919311e+01, 9.81920114e+00, 3.62142857e-02,\n","       2.28809524e-02, 2.76185941e-03, 2.81056704e+01, 2.04208025e+02,\n","       7.29051694e+02, 2.00572701e+03, 4.21366647e+03, 6.92346357e+03,\n","       9.57389827e+03, 1.14603398e+04, 1.24666788e+04, 1.27682484e+04,\n","       1.26613017e+04, 1.25534327e+04, 1.25998206e+04, 1.23927837e+04,\n","       1.15590439e+04, 9.88953142e+03, 7.53295513e+03, 5.11441259e+03,\n","       3.05547189e+03, 1.66092585e+03, 8.39969863e+02, 3.84512312e+02,\n","       1.34737119e+02, 2.71408189e+01, 3.06857624e+00, 1.23428571e-01,\n","       0.00000000e+00, 0.00000000e+00, 7.31084788e+00, 7.57570670e+01,\n","       3.18535551e+02, 8.92561149e+02, 2.13007164e+03, 4.15513729e+03,\n","       6.67943441e+03, 9.08942045e+03, 1.08768840e+04, 1.18600478e+04,\n","       1.22185140e+04, 1.21488922e+04, 1.16761724e+04, 1.06623427e+04,\n","       8.88180421e+03, 6.72867491e+03, 4.52696553e+03, 2.78125867e+03,\n","       1.57340674e+03, 7.99627345e+02, 3.90776811e+02, 1.74143994e+02,\n","       5.13083781e+01, 8.53411049e+00, 1.14082626e+00, 0.00000000e+00,\n","       0.00000000e+00, 0.00000000e+00, 3.09874806e+00, 2.12014616e+01,\n","       9.01086432e+01, 2.89578069e+02, 7.37002996e+02, 1.63305438e+03,\n","       3.00614325e+03, 4.75217719e+03, 6.42418657e+03, 7.68266614e+03,\n","       8.23705320e+03, 8.21603450e+03, 7.48615583e+03, 6.25364040e+03,\n","       4.88502834e+03, 3.50985901e+03, 2.25424437e+03, 1.40076179e+03,\n","       7.54457368e+02, 3.85015891e+02, 1.70931569e+02, 7.23728862e+01,\n","       1.53588609e+01, 3.76286683e+00, 2.36378952e-01, 0.00000000e+00,\n","       0.00000000e+00, 0.00000000e+00, 7.89492516e-02, 3.23002085e+00,\n","       2.66993369e+01, 8.62829380e+01, 2.42567552e+02, 5.74670736e+02,\n","       1.16062824e+03, 1.85908331e+03, 2.63430007e+03, 3.18296192e+03,\n","       3.47208469e+03, 3.42498298e+03, 3.04423070e+03, 2.59463417e+03,\n","       2.08583729e+03, 1.54170230e+03, 1.04262030e+03, 6.17155891e+02,\n","       3.32872699e+02, 1.63141373e+02, 6.60019146e+01, 2.49646484e+01,\n","       2.80197112e+00, 5.64225568e-01, 2.57523810e-01, 0.00000000e+00,\n","       0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,\n","       5.85748256e+00, 2.32008075e+01, 9.24954784e+01, 2.19439668e+02,\n","       4.27698080e+02, 6.31131067e+02, 9.05773738e+02, 1.16529687e+03,\n","       1.28219277e+03, 1.23351933e+03, 1.09034012e+03, 8.82789099e+02,\n","       7.26724558e+02, 5.23459986e+02, 3.45778063e+02, 2.08353021e+02,\n","       1.10624593e+02, 4.18520306e+01, 1.58110107e+01, 3.40777397e+00,\n","       1.94759195e-02, 8.28809524e-02, 0.00000000e+00, 0.00000000e+00,\n","       0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,\n","       0.00000000e+00, 9.02126602e-01, 5.52653803e+00, 1.54785620e+01,\n","       2.06441456e+01, 3.43017827e+01, 5.21246770e+01, 7.97142975e+01,\n","       1.00081387e+02, 1.02608696e+02, 1.26673478e+02, 1.14405059e+02,\n","       8.98716430e+01, 6.32064973e+01, 3.98525769e+01, 2.14722761e+01,\n","       1.07222715e+01, 3.09714228e+00, 3.58912164e+00, 1.71614970e-01,\n","       0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00])"]},"metadata":{"tags":[]},"execution_count":9}]},{"metadata":{"id":"AICU6nOqnvOC","colab_type":"code","colab":{}},"cell_type":"code","source":["index = xx.argsort()"],"execution_count":0,"outputs":[]},{"metadata":{"id":"0G4RsXSKrRsX","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":1071},"outputId":"802540ce-4ca6-4eab-d4c3-a89cda1173c4","executionInfo":{"status":"ok","timestamp":1546132468008,"user_tz":-480,"elapsed":788,"user":{"displayName":"Sen Yang","photoUrl":"","userId":"00832503676208839570"}}},"cell_type":"code","source":["index"],"execution_count":16,"outputs":[{"output_type":"execute_result","data":{"text/plain":["array([  0, 448, 421, 420, 392, 782, 196, 168, 141, 140, 476, 139, 111,\n","        85,  84,  83,  82,  57,  56,  55,  54, 112, 532, 560, 644, 781,\n","       780, 760, 759, 758, 757, 756, 755, 754, 731, 730, 729, 728, 727,\n","       701, 700, 699, 673, 672, 671, 645,  53,  52, 783,   7,  28,  18,\n","        29,  19,  30,  20,  27,  11,  10,  26,  21,   9,   8,  17,  16,\n","         5,  22,  31,  23,   4,   3,  25,   2,   1,  24,   6, 169,  15,\n","       617,  32, 167, 752,  59, 616, 588, 364, 113, 615,  33, 587, 224,\n","       449, 702, 419, 753,  58, 114, 643, 504, 779, 477, 698, 726,  87,\n","        12, 110, 505,  86, 589, 725,  61, 559, 336, 308, 142, 335,  60,\n","       393, 761,  51, 363,  81,  14, 670, 561, 115,  34, 447,  88, 531,\n","       252, 197, 503,  50, 533, 365, 109, 251, 170,  49,  13, 223, 724,\n","       475, 391, 642, 777, 674, 280, 703, 751, 195, 778, 697, 422, 307,\n","       279, 143,  35, 138, 762, 732, 337,  80, 646, 669,  48, 394, 418,\n","        62,  36, 614, 450,  37, 776, 309, 225, 116, 390,  47, 362,  89,\n","       696, 763, 586, 750,  79, 198, 478, 166, 253, 281,  46, 764, 334,\n","       675, 775, 733, 366,  38, 723, 530, 446, 704, 641, 506, 618, 558,\n","        63,  45, 137,  44, 474, 108, 502,  43,  39, 765,  41,  40,  42,\n","       306, 171, 774, 749,  90, 590, 534, 338, 668, 278, 766, 194,  78,\n","       562,  64, 144, 117, 773, 722, 226, 695, 613, 395, 361, 647, 250,\n","       222, 107, 767, 310, 423, 705, 772, 676, 734,  91, 768, 282,  65,\n","       254, 389, 769, 199, 333, 748, 451, 771,  77, 367, 165, 770, 585,\n","       640, 417, 136, 721,  66, 479, 694, 118, 339, 667, 557, 445, 106,\n","       305, 619, 172, 747,  92, 735, 529, 473, 501,  76, 706, 311,  67,\n","       145, 227, 507, 677, 612, 193, 277, 648, 720, 283, 746, 591, 255,\n","        68, 535, 639, 693, 135, 666,  75, 221, 736, 563, 119, 249,  93,\n","       105, 200, 584, 164,  69,  74, 368, 745, 396, 340, 707, 173, 719,\n","       424, 737, 556,  73,  70, 146, 312,  72,  71,  94, 744, 620, 678,\n","       332, 692, 228, 611, 104, 665, 452, 360, 528, 304, 134, 638, 284,\n","       743, 649, 738, 120, 256, 500, 192, 388, 718, 472, 276, 742, 480,\n","        95, 416, 201, 444, 163, 708, 739, 220, 583, 741, 592, 248, 740,\n","       103, 147, 174, 691, 508,  96, 717, 664, 610, 133, 121, 679, 637,\n","       229, 564, 536, 555, 313, 341, 102, 709, 285, 257,  97, 369, 621,\n","       191, 716, 527, 650, 101, 162, 690, 397, 202,  98, 331, 148, 499,\n","       303, 100, 359, 582, 122,  99, 715, 710, 175, 425, 219, 471, 387,\n","       132, 275, 663, 443, 415, 247, 609, 680, 714, 636, 453, 230, 711,\n","       593, 713, 481, 712, 689, 554, 258, 123, 190, 286, 509, 161, 314,\n","       149, 565, 203, 537, 131, 651, 526, 342, 622, 176, 662, 358, 386,\n","       581, 498, 370, 330, 681, 414, 688, 218, 124, 470, 442, 635, 302,\n","       608, 398, 231, 130, 274, 246, 426, 150, 594, 160, 454, 553, 189,\n","       482, 259, 125, 687, 204, 510, 682, 287, 177, 385, 129, 566, 357,\n","       538, 652, 525, 661, 315, 413, 623, 580, 126, 497, 441, 329, 128,\n","       469, 343, 686, 151, 634, 127, 217, 683, 607, 232, 159, 371, 301,\n","       483, 511, 685, 684, 455, 399, 245, 552, 273, 427, 539, 595, 205,\n","       178, 188, 384, 260, 412, 660, 567, 356, 653, 524, 440, 152, 512,\n","       288, 496, 468, 484, 624, 579, 540, 158, 328, 316, 633, 321, 513,\n","       606, 233, 216, 456, 485, 344, 179, 206, 300, 568, 322, 514, 541,\n","       294, 428, 551, 659, 486, 596, 349, 187, 153, 372, 293, 320, 348,\n","       400, 411, 654, 244, 439, 272, 467, 383, 515, 261, 295, 157, 495,\n","       457, 487, 523, 542, 355, 458, 516, 323, 376, 569, 578, 543, 289,\n","       625, 296, 459, 154, 632, 207, 292, 234, 327, 605, 429, 180, 404,\n","       156, 658, 324, 317, 403, 494, 431, 347, 375, 430, 466, 215, 488,\n","       544, 377, 267, 155, 266, 319, 655, 350, 597, 570, 517, 268, 299,\n","       345, 401, 522, 550, 373, 402, 435, 407, 408, 432, 186, 297, 460,\n","       438, 265, 208, 493, 271, 352, 380, 325, 545, 211, 521, 212, 243,\n","       374, 657, 571, 520, 601, 262, 326, 318, 269, 346, 656, 210, 602,\n","       209, 298, 213, 548, 290, 604, 291, 577, 240, 235, 214, 489, 239,\n","       351, 603, 241, 519, 518, 354, 492, 598, 572, 600, 549, 353, 181,\n","       236, 238, 264, 547, 463, 576, 465, 631, 575, 410, 270, 436, 574,\n","       573, 242, 599, 405, 546, 237, 379, 626, 263, 382, 491, 490, 629,\n","       185, 381, 182, 630, 464, 183, 184, 409, 628, 437, 462, 433, 434,\n","       461, 627, 378, 406])"]},"metadata":{"tags":[]},"execution_count":16}]},{"metadata":{"id":"sIpgnbuYoKDb","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":35},"outputId":"8c5d33e3-d6ae-48c0-a017-061264706325","executionInfo":{"status":"ok","timestamp":1546132525882,"user_tz":-480,"elapsed":645,"user":{"displayName":"Sen Yang","photoUrl":"","userId":"00832503676208839570"}}},"cell_type":"code","source":["th = xx[index[-62]]\n","th"],"execution_count":20,"outputs":[{"output_type":"execute_result","data":{"text/plain":["12279.38110199273"]},"metadata":{"tags":[]},"execution_count":20}]},{"metadata":{"id":"E82rrQuHnn6w","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":35},"outputId":"555f5fed-6f18-45ce-d2ba-cdeb2ef1e4da","executionInfo":{"status":"ok","timestamp":1546132528607,"user_tz":-480,"elapsed":915,"user":{"displayName":"Sen Yang","photoUrl":"","userId":"00832503676208839570"}}},"cell_type":"code","source":["from sklearn.feature_selection import VarianceThreshold\n","X_fsvar = VarianceThreshold(th).fit_transform(X)\n","X_fsvar.shape"],"execution_count":21,"outputs":[{"output_type":"execute_result","data":{"text/plain":["(42000, 61)"]},"metadata":{"tags":[]},"execution_count":21}]},{"metadata":{"id":"vwi6-KgfoYuU","colab_type":"text"},"cell_type":"markdown","source":["包装法"]},{"metadata":{"id":"ikxzclOpob3a","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":35},"outputId":"df911302-4ffd-4594-ea39-d0ee1007f429","executionInfo":{"status":"ok","timestamp":1546132614951,"user_tz":-480,"elapsed":61316,"user":{"displayName":"Sen Yang","photoUrl":"","userId":"00832503676208839570"}}},"cell_type":"code","source":["from sklearn.feature_selection import RFE\n","selector = RFE(RFC_, n_features_to_select=61, step=50).fit(X, y)\n","X_wrapper = selector.transform(X)\n","X_wrapper.shape"],"execution_count":22,"outputs":[{"output_type":"execute_result","data":{"text/plain":["(42000, 61)"]},"metadata":{"tags":[]},"execution_count":22}]},{"metadata":{"id":"pH9iQTqBstc1","colab_type":"text"},"cell_type":"markdown","source":["## 结果对比"]},{"metadata":{"id":"CGA5nERPovHW","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":35},"outputId":"59424cf1-7543-4f94-821b-7d80c7c03f44","executionInfo":{"status":"ok","timestamp":1546132663941,"user_tz":-480,"elapsed":8328,"user":{"displayName":"Sen Yang","photoUrl":"","userId":"00832503676208839570"}}},"cell_type":"code","source":["from sklearn.model_selection import cross_val_score\n","cross_val_score(RFC_,X_fsvar,y,cv=5).mean()"],"execution_count":24,"outputs":[{"output_type":"execute_result","data":{"text/plain":["0.8599997381364645"]},"metadata":{"tags":[]},"execution_count":24}]},{"metadata":{"id":"sAsY5tzIovO3","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":35},"outputId":"4ad473e4-d8f7-4ab9-e893-aaf3d7fb432d","executionInfo":{"status":"ok","timestamp":1546132693980,"user_tz":-480,"elapsed":7663,"user":{"displayName":"Sen Yang","photoUrl":"","userId":"00832503676208839570"}}},"cell_type":"code","source":["cross_val_score(RFC_,X_embedded,y,cv=5).mean()"],"execution_count":25,"outputs":[{"output_type":"execute_result","data":{"text/plain":["0.9083817360504339"]},"metadata":{"tags":[]},"execution_count":25}]},{"metadata":{"id":"l5IxAx2WolJ0","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":35},"outputId":"3a1c78ab-4401-48fc-d76f-b115f24ed88a","executionInfo":{"status":"ok","timestamp":1546132704071,"user_tz":-480,"elapsed":7776,"user":{"displayName":"Sen Yang","photoUrl":"","userId":"00832503676208839570"}}},"cell_type":"code","source":["cross_val_score(RFC_,X_wrapper,y,cv=5).mean()"],"execution_count":26,"outputs":[{"output_type":"execute_result","data":{"text/plain":["0.9081908885794439"]},"metadata":{"tags":[]},"execution_count":26}]},{"metadata":{"id":"Fj9YAbrfsKyM","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":35},"outputId":"48ec26db-2b1e-49d6-cf95-7729359334df","executionInfo":{"status":"ok","timestamp":1546132816758,"user_tz":-480,"elapsed":68613,"user":{"displayName":"Sen Yang","photoUrl":"","userId":"00832503676208839570"}}},"cell_type":"code","source":["cross_val_score(RFC(n_estimators=100,random_state=0),X_embedded,y,cv=5).mean()"],"execution_count":27,"outputs":[{"output_type":"execute_result","data":{"text/plain":["0.9338340363576279"]},"metadata":{"tags":[]},"execution_count":27}]},{"metadata":{"id":"TZ_FgPIpsXcH","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":35},"outputId":"69e536cd-9ca4-4676-fd3f-5ee7b7dd4a71","executionInfo":{"status":"ok","timestamp":1546132974258,"user_tz":-480,"elapsed":68151,"user":{"displayName":"Sen Yang","photoUrl":"","userId":"00832503676208839570"}}},"cell_type":"code","source":["cross_val_score(RFC(n_estimators=100,random_state=0),X_wrapper,y,cv=5).mean()"],"execution_count":29,"outputs":[{"output_type":"execute_result","data":{"text/plain":["0.933310464765106"]},"metadata":{"tags":[]},"execution_count":29}]},{"metadata":{"id":"OdVfVI-XsbKv","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":35},"outputId":"7e5998c2-5898-4931-f301-5646e1ff828c","executionInfo":{"status":"ok","timestamp":1546132898326,"user_tz":-480,"elapsed":75108,"user":{"displayName":"Sen Yang","photoUrl":"","userId":"00832503676208839570"}}},"cell_type":"code","source":["cross_val_score(RFC(n_estimators=100,random_state=0),X_fsvar,y,cv=5).mean()"],"execution_count":28,"outputs":[{"output_type":"execute_result","data":{"text/plain":["0.8968338541459806"]},"metadata":{"tags":[]},"execution_count":28}]},{"metadata":{"id":"-hQOfIpuuAYq","colab_type":"text"},"cell_type":"markdown","source":["## 特征更少时的对比"]},{"metadata":{"id":"v_0u1aScspxD","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":35},"outputId":"4960bf62-3f5a-4706-96e1-8ff28f566251","executionInfo":{"status":"ok","timestamp":1546133235302,"user_tz":-480,"elapsed":4624,"user":{"displayName":"Sen Yang","photoUrl":"","userId":"00832503676208839570"}}},"cell_type":"code","source":["from sklearn.feature_selection import SelectFromModel\n","from sklearn.ensemble import RandomForestClassifier as RFC\n","RFC_ = RFC(n_estimators =10,random_state=0)\n","X_embedded2 = SelectFromModel(RFC_,threshold=0.005).fit_transform(X,y)\n","X_embedded2.shape\n"],"execution_count":30,"outputs":[{"output_type":"execute_result","data":{"text/plain":["(42000, 47)"]},"metadata":{"tags":[]},"execution_count":30}]},{"metadata":{"id":"kFib1X4NuNPS","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":35},"outputId":"5c162c51-8d75-4e44-c9dc-1d5855458def","executionInfo":{"status":"ok","timestamp":1546133323195,"user_tz":-480,"elapsed":60158,"user":{"displayName":"Sen Yang","photoUrl":"","userId":"00832503676208839570"}}},"cell_type":"code","source":["from sklearn.feature_selection import RFE\n","selector2 = RFE(RFC_, n_features_to_select=47, step=50).fit(X, y)\n","X_wrapper2 = selector2.transform(X)\n","X_wrapper2.shape"],"execution_count":31,"outputs":[{"output_type":"execute_result","data":{"text/plain":["(42000, 47)"]},"metadata":{"tags":[]},"execution_count":31}]},{"metadata":{"id":"VWH5BitauVIm","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":35},"outputId":"1b1571ea-c587-4f0a-bb0f-4323e80e4481","executionInfo":{"status":"ok","timestamp":1546133366267,"user_tz":-480,"elapsed":6761,"user":{"displayName":"Sen Yang","photoUrl":"","userId":"00832503676208839570"}}},"cell_type":"code","source":["cross_val_score(RFC_,X_embedded2,y,cv=5).mean()"],"execution_count":34,"outputs":[{"output_type":"execute_result","data":{"text/plain":["0.8972375307515474"]},"metadata":{"tags":[]},"execution_count":34}]},{"metadata":{"id":"uTFtYhJ5uZmp","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":35},"outputId":"ce5c6c5e-c5bf-41e8-c218-7de5b749c2ec","executionInfo":{"status":"ok","timestamp":1546133375837,"user_tz":-480,"elapsed":6679,"user":{"displayName":"Sen Yang","photoUrl":"","userId":"00832503676208839570"}}},"cell_type":"code","source":["cross_val_score(RFC_,X_wrapper2,y,cv=5).mean()"],"execution_count":35,"outputs":[{"output_type":"execute_result","data":{"text/plain":["0.9020955986904144"]},"metadata":{"tags":[]},"execution_count":35}]},{"metadata":{"id":"6k_IX0xFuqID","colab_type":"code","colab":{}},"cell_type":"code","source":[""],"execution_count":0,"outputs":[]}]}

一键复制编辑原始数据按行查看历史

提交于 2019-07-10 18:55 +08:00 . first

Loading...

马建仓 AI 助手

尝试更多

代码解读

代码找茬

代码优化

1

https://gitee.com/likunzm/sklearn-ml.git

git@gitee.com:likunzm/sklearn-ml.git

likunzm

sklearn-ml

sklearn-ml

master