File size: 4,176 Bytes
d8e0581 a68744f d8e0581 a68744f d8e0581 a68744f d8e0581 a68744f d8e0581 95c1d69 8a32f4d 53b80a9 8a32f4d 95c1d69 d8e0581 95c1d69 a68744f d8e0581 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 |
import gradio as gr
import numpy as np
from sklearn.datasets import load_diabetes
from sklearn.linear_model import RidgeCV
from sklearn.feature_selection import SelectFromModel
from time import time
from sklearn.feature_selection import SequentialFeatureSelector
import matplotlib
matplotlib.use("Agg")
import matplotlib.pyplot as plt
def select_features(method,num_features):
diabetes = load_diabetes()
X, y = diabetes.data, diabetes.target
ridge = RidgeCV(alphas=np.logspace(-6, 6, num=5)).fit(X, y)
feature_names = np.array(diabetes.feature_names)
if method == 'model':
importance = np.abs(ridge.coef_)
tic = time()
sfm = SelectFromModel(ridge, threshold=-np.inf,max_features=num_features).fit(X, y)
toc = time()
selected_features = feature_names[sfm.get_support()]
if int(num_features) < len(selected_features):
selected_features = selected_features[:int(num_features)]
execution_time = toc - tic
fig, ax = plt.subplots()
ax.bar(height=importance, x=feature_names)
ax.set_title("Feature importances via coefficients")
ax.set_ylabel("Importance coefficient")
ax.set_xlabel("Features")
elif method == 'sfs-forward':
tic_fwd = time()
sfs_forward = SequentialFeatureSelector(
ridge, n_features_to_select=int(num_features), direction="forward"
).fit(X, y)
toc_fwd = time()
selected_features = feature_names[sfs_forward.get_support()]
execution_time = toc_fwd - tic_fwd
importance = np.abs(sfs_forward.get_params()['estimator'].coef_)
fig = None
elif method == 'sfs-backward':
tic_bwd = time()
sfs_backward = SequentialFeatureSelector(
ridge, n_features_to_select=int(num_features), direction="backward"
).fit(X, y)
toc_bwd = time()
selected_features = feature_names[sfs_backward.get_support()]
execution_time = toc_bwd - tic_bwd
importance = np.abs(sfs_backward.get_params()['estimator'].coef_)
fig = None
return f"Selected the following features: {', '.join(selected_features)} in {execution_time:.3f} seconds", fig
title = "Selecting features with Sequential Feature Selection"
with gr.Blocks(title=title) as demo:
gr.Markdown(f"## {title}")
gr.Markdown("""
This app demonstrates feature selection techniques using model based selection and sequential feature selection.\n\n
Model based selection is based on feature importance. Each feature is assigned a score on how much influence they have on the model output.
The feature with highest score is considered the most important feature.\n\n
Sequential feature selection is based on greedy approach. In greedy approach, the feature is added or removed to the selected features at each iteration
based on the model performance score.\n\n
This app uses Ridge estimator and the diabetes dataset from sklearn. Diabetes dataset consist of quantitative measure of diabetes progression and
10 following variables obtained from 442 diabetes patients:
1. Age (age)
2. Sex (sex)
3. Body mass index (bmi)
4. Average blood pressure (bp)
5. Total serum cholesterol (s1)
6. Low-density lipoproteins (s2)
7. High-density lipoproteins (s3)
8. Total cholesterol / HDL (s4)
9. Possibly log of serum triglycerides level (s5)
10. Blood sugar level (s6)\n\n
This app is developed based on [scikit-learn example](https://scikit-learn.org/stable/auto_examples/feature_selection/plot_select_from_model_diabetes.html#sphx-glr-auto-examples-feature-selection-plot-select-from-model-diabetes-py)
""")
method = gr.Radio(["model", "sfs-forward", "sfs-backward"], label="Method")
num_features = gr.Slider(minimum=2, maximum=10, step=1, label = "Number of features")
output = gr.Textbox(label="Selected features")
plot = gr.Plot(label="Feature importance plot")
num_features.change(fn=select_features, inputs=[method,num_features], outputs=[output,plot])
demo.launch()
|