"
]
},
"metadata": {
"tags": [],
"needs_background": "light"
}
}
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "x7kKRuulE72M",
"colab_type": "text"
},
"source": [
"This is a highly imablanced dataset."
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "hGF3OuZMq1ra",
"colab_type": "text"
},
"source": [
"## Problem with imbalanced dataset: \n",
"We need to deal with the dataset in a correct way. When we will train our model than our model will achieve high accuracy but our trained model will predict a negative class in maximum number of cases. So, we also need to keep in mind the precision and recall score in such scenarios. \n",
"\n",
"This problem is predominant in scenarios where anomaly detection is crucial like electricity pilferage, fraudulent transactions in banks, identification of rare diseases, etc. In this situation, the predictive model developed using conventional machine learning algorithms could be biased and inaccurate.\n",
"\n",
"This happens because Machine Learning Algorithms are usually designed to improve accuracy by reducing the error. Thus, they do not take into account the class distribution / proportion or balance of classes.\n",
"\n",
"This guide describes various approaches for solving such class imbalance problems using Pycaret. "
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "Uch03o6zsn-h",
"colab_type": "text"
},
"source": [
"## Prepairing the setup"
]
},
{
"cell_type": "code",
"metadata": {
"id": "_gxUecGVsnAM",
"colab_type": "code",
"colab": {}
},
"source": [
"from pycaret.classification import *\n"
],
"execution_count": 8,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "0MDZGQo6p9bC",
"colab_type": "code",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 918,
"referenced_widgets": [
"0955e5b8be8042bbae4ee23c230a61eb",
"7d901db567184b758cd7e56ab776f1d3",
"9169f9e763314fddb72dfea8cabd09a4",
"727e71d800a742bcb3c81ed2e8e964ee",
"f9a6f7e7b553446392ca730a47f4bb3a",
"a84f1abc70714e33a58147e1c6f995ef"
]
},
"outputId": "32779414-a475-48a6-bba2-df9d97d9551f"
},
"source": [
"clf=setup(data=df,target='Class',fix_imbalance=True) #fix_imbalance will automaticaaly fix the imbalanced dataset by oversampling using the SMOTE method."
],
"execution_count": 9,
"outputs": [
{
"output_type": "stream",
"text": [
"Setup Succesfully Completed!\n"
],
"name": "stdout"
},
{
"output_type": "display_data",
"data": {
"text/html": [
"
Description
Value
\n",
"
\n",
"
0
\n",
"
session_id
\n",
"
3458
\n",
"
\n",
"
\n",
"
1
\n",
"
Target Type
\n",
"
Binary
\n",
"
\n",
"
\n",
"
2
\n",
"
Label Encoded
\n",
"
None
\n",
"
\n",
"
\n",
"
3
\n",
"
Original Data
\n",
"
(284807, 31)
\n",
"
\n",
"
\n",
"
4
\n",
"
Missing Values
\n",
"
False
\n",
"
\n",
"
\n",
"
5
\n",
"
Numeric Features
\n",
"
30
\n",
"
\n",
"
\n",
"
6
\n",
"
Categorical Features
\n",
"
0
\n",
"
\n",
"
\n",
"
7
\n",
"
Ordinal Features
\n",
"
False
\n",
"
\n",
"
\n",
"
8
\n",
"
High Cardinality Features
\n",
"
False
\n",
"
\n",
"
\n",
"
9
\n",
"
High Cardinality Method
\n",
"
None
\n",
"
\n",
"
\n",
"
10
\n",
"
Sampled Data
\n",
"
(199364, 31)
\n",
"
\n",
"
\n",
"
11
\n",
"
Transformed Train Set
\n",
"
(139554, 30)
\n",
"
\n",
"
\n",
"
12
\n",
"
Transformed Test Set
\n",
"
(59810, 30)
\n",
"
\n",
"
\n",
"
13
\n",
"
Numeric Imputer
\n",
"
mean
\n",
"
\n",
"
\n",
"
14
\n",
"
Categorical Imputer
\n",
"
constant
\n",
"
\n",
"
\n",
"
15
\n",
"
Normalize
\n",
"
False
\n",
"
\n",
"
\n",
"
16
\n",
"
Normalize Method
\n",
"
None
\n",
"
\n",
"
\n",
"
17
\n",
"
Transformation
\n",
"
False
\n",
"
\n",
"
\n",
"
18
\n",
"
Transformation Method
\n",
"
None
\n",
"
\n",
"
\n",
"
19
\n",
"
PCA
\n",
"
False
\n",
"
\n",
"
\n",
"
20
\n",
"
PCA Method
\n",
"
None
\n",
"
\n",
"
\n",
"
21
\n",
"
PCA Components
\n",
"
None
\n",
"
\n",
"
\n",
"
22
\n",
"
Ignore Low Variance
\n",
"
False
\n",
"
\n",
"
\n",
"
23
\n",
"
Combine Rare Levels
\n",
"
False
\n",
"
\n",
"
\n",
"
24
\n",
"
Rare Level Threshold
\n",
"
None
\n",
"
\n",
"
\n",
"
25
\n",
"
Numeric Binning
\n",
"
False
\n",
"
\n",
"
\n",
"
26
\n",
"
Remove Outliers
\n",
"
False
\n",
"
\n",
"
\n",
"
27
\n",
"
Outliers Threshold
\n",
"
None
\n",
"
\n",
"
\n",
"
28
\n",
"
Remove Multicollinearity
\n",
"
False
\n",
"
\n",
"
\n",
"
29
\n",
"
Multicollinearity Threshold
\n",
"
None
\n",
"
\n",
"
\n",
"
30
\n",
"
Clustering
\n",
"
False
\n",
"
\n",
"
\n",
"
31
\n",
"
Clustering Iteration
\n",
"
None
\n",
"
\n",
"
\n",
"
32
\n",
"
Polynomial Features
\n",
"
False
\n",
"
\n",
"
\n",
"
33
\n",
"
Polynomial Degree
\n",
"
None
\n",
"
\n",
"
\n",
"
34
\n",
"
Trignometry Features
\n",
"
False
\n",
"
\n",
"
\n",
"
35
\n",
"
Polynomial Threshold
\n",
"
None
\n",
"
\n",
"
\n",
"
36
\n",
"
Group Features
\n",
"
False
\n",
"
\n",
"
\n",
"
37
\n",
"
Feature Selection
\n",
"
False
\n",
"
\n",
"
\n",
"
38
\n",
"
Features Selection Threshold
\n",
"
None
\n",
"
\n",
"
\n",
"
39
\n",
"
Feature Interaction
\n",
"
False
\n",
"
\n",
"
\n",
"
40
\n",
"
Feature Ratio
\n",
"
False
\n",
"
\n",
"
\n",
"
41
\n",
"
Interaction Threshold
\n",
"
None
\n",
"
\n",
"
\n",
"
42
\n",
"
Fix Imbalance
\n",
"
True
\n",
"
\n",
"
\n",
"
43
\n",
"
Fix Imbalance Method
\n",
"
SMOTE
\n",
"
\n",
"
"
],
"text/plain": [
""
]
},
"metadata": {
"tags": []
}
}
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "nQyNyXctwAGx",
"colab_type": "text"
},
"source": [
"### SMOTE method: SMOTe is a technique based on nearest neighbors judged by Euclidean Distance between data points in feature space. There is a percentage of Over-Sampling which indicates the number of synthetic samples to be created and this percentage parameter of Over-sampling is always a multiple of 100."
]
},
{
"cell_type": "code",
"metadata": {
"id": "SU3zRpKcVgi0",
"colab_type": "code",
"colab": {}
},
"source": [
"#Uncomment the following code to compare the performance of all the classification models\n",
"\n",
"\n",
"#compare_models() "
],
"execution_count": null,
"outputs": []
},
{
"cell_type": "markdown",
"metadata": {
"id": "PBAqx-IuPGb5",
"colab_type": "text"
},
"source": [
"## We will choose the model with high precision because here we need to have a high precision than high accuracy or high recalls.\n",
"\n",
"This link will provide you some overview of precision and recall.\n",
"Link: https://developers.google.com/machine-learning/crash-course/classification/precision-and-recall"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "wNdaW3J5V3rk",
"colab_type": "text"
},
"source": [
"We are creating the random forest classifier because it works really well with these types of dataset. You can have a quick view of the different models using 'compare_models()'"
]
},
{
"cell_type": "code",
"metadata": {
"id": "0RHd8qp1wL9Y",
"colab_type": "code",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 407,
"referenced_widgets": [
"ac8d355c4fd04396a4b3962e69b57e78",
"e359da09f13c44939189b6dd7a8f7c7f",
"bb8c34bfea704c3ca6f5733b4ed15f4d"
]
},
"outputId": "5571d088-6ef1-47fa-9377-a4f09258be91"
},
"source": [
"classifier=create_model('rf')\n",
"print(classifier)"
],
"execution_count": 12,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/html": [
"
Accuracy
AUC
Recall
Prec.
F1
Kappa
MCC
\n",
"
\n",
"
0
\n",
"
0.9996
\n",
"
0.9782
\n",
"
0.8333
\n",
"
0.9524
\n",
"
0.8889
\n",
"
0.8887
\n",
"
0.8907
\n",
"
\n",
"
\n",
"
1
\n",
"
0.9994
\n",
"
0.8722
\n",
"
0.7083
\n",
"
0.8947
\n",
"
0.7907
\n",
"
0.7904
\n",
"
0.7958
\n",
"
\n",
"
\n",
"
2
\n",
"
0.9994
\n",
"
0.9572
\n",
"
0.8750
\n",
"
0.8077
\n",
"
0.8400
\n",
"
0.8397
\n",
"
0.8404
\n",
"
\n",
"
\n",
"
3
\n",
"
0.9997
\n",
"
0.9386
\n",
"
0.8800
\n",
"
0.9565
\n",
"
0.9167
\n",
"
0.9165
\n",
"
0.9173
\n",
"
\n",
"
\n",
"
4
\n",
"
0.9994
\n",
"
0.9147
\n",
"
0.7500
\n",
"
0.8571
\n",
"
0.8000
\n",
"
0.7997
\n",
"
0.8015
\n",
"
\n",
"
\n",
"
5
\n",
"
0.9997
\n",
"
0.9574
\n",
"
0.9167
\n",
"
0.9167
\n",
"
0.9167
\n",
"
0.9165
\n",
"
0.9165
\n",
"
\n",
"
\n",
"
6
\n",
"
0.9995
\n",
"
0.9142
\n",
"
0.7917
\n",
"
0.9048
\n",
"
0.8444
\n",
"
0.8442
\n",
"
0.8461
\n",
"
\n",
"
\n",
"
7
\n",
"
0.9994
\n",
"
0.9141
\n",
"
0.7917
\n",
"
0.8261
\n",
"
0.8085
\n",
"
0.8082
\n",
"
0.8084
\n",
"
\n",
"
\n",
"
8
\n",
"
0.9994
\n",
"
0.9359
\n",
"
0.8333
\n",
"
0.8333
\n",
"
0.8333
\n",
"
0.8330
\n",
"
0.8330
\n",
"
\n",
"
\n",
"
9
\n",
"
0.9995
\n",
"
0.9358
\n",
"
0.7500
\n",
"
0.9474
\n",
"
0.8372
\n",
"
0.8370
\n",
"
0.8427
\n",
"
\n",
"
\n",
"
Mean
\n",
"
0.9995
\n",
"
0.9318
\n",
"
0.8130
\n",
"
0.8897
\n",
"
0.8476
\n",
"
0.8474
\n",
"
0.8492
\n",
"
\n",
"
\n",
"
SD
\n",
"
0.0001
\n",
"
0.0283
\n",
"
0.0630
\n",
"
0.0526
\n",
"
0.0432
\n",
"
0.0433
\n",
"
0.0425
\n",
"
\n",
"
"
],
"text/plain": [
""
]
},
"metadata": {
"tags": []
}
},
{
"output_type": "stream",
"text": [
"RandomForestClassifier(bootstrap=True, ccp_alpha=0.0, class_weight=None,\n",
" criterion='gini', max_depth=None, max_features='auto',\n",
" max_leaf_nodes=None, max_samples=None,\n",
" min_impurity_decrease=0.0, min_impurity_split=None,\n",
" min_samples_leaf=1, min_samples_split=2,\n",
" min_weight_fraction_leaf=0.0, n_estimators=10, n_jobs=-1,\n",
" oob_score=False, random_state=3458, verbose=0,\n",
" warm_start=False)\n"
],
"name": "stdout"
}
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "aOH_EldwQ5Rt",
"colab_type": "text"
},
"source": [
"## Classification plots"
]
},
{
"cell_type": "code",
"metadata": {
"id": "kjbCMz7KQ43s",
"colab_type": "code",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 401,
"referenced_widgets": [
"b974713bf81244b8ad8567b1b797b3a0",
"1cc5d7149d994632a82330b094bd257f",
"776d2b0cd8fc416790be305938018d61"
]
},
"outputId": "ed8ab10d-6a8a-4e9e-8f2c-234b2d5c9548"
},
"source": [
"# Plotting the classification report\n",
"plot_model(classifier,plot='class_report')\n"
],
"execution_count": 16,
"outputs": [
{
"output_type": "display_data",
"data": {
"image/png": "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\n",
"text/plain": [
""
]
},
"metadata": {
"tags": []
}
}
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "8ib3-Oc7X_6s",
"colab_type": "text"
},
"source": [
"### Here important point to notice is the precision, recall, and f1 score for the positive class that is '1'"
]
},
{
"cell_type": "code",
"metadata": {
"id": "ZdE1WS5cQ4q9",
"colab_type": "code",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 374,
"referenced_widgets": [
"33b14c76463243c89e55d95509cab12d"
]
},
"outputId": "99306c54-e58b-40d3-c1b0-fe6a0956f314"
},
"source": [
"# Plotting the confusion matrix\n",
"plot_model(classifier,plot='confusion_matrix')\n"
],
"execution_count": 17,
"outputs": [
{
"output_type": "display_data",
"data": {
"image/png": "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\n",
"text/plain": [
""
]
},
"metadata": {
"tags": []
}
}
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "BPKk8BRnRWcO",
"colab_type": "text"
},
"source": [
"## Now, from the above plots we can easily conclude that we succesfully handled our highly imbalanced dataset as our precision score is high(90.1%). We would got a high accuracy score and recall score but a very low precision score, if we haven't succesfully handled our imbalanced dataset."
]
},
{
"cell_type": "code",
"metadata": {
"id": "vx-bh4yQXG7x",
"colab_type": "code",
"colab": {}
},
"source": [
""
],
"execution_count": null,
"outputs": []
}
]
}