1 . Data Pre-Processing 2 . Model Building
This project doesn't required too much data pre-processing technique because it mainly focuses on applying different machine learning algorithms and selecting which algorithms is best suited for my dataset Data Preprocessing is maily consist of two parts 1 . Splitting Data into test and train 2 . Data Normalisation and Standardization
I have applied different classification algorithms which are stated as below with observed accuracy 1 . Logistic Regression
Algorithms | Type | Accuracy |
---|---|---|
Logistic Regression | Simple | 83.0% |
Logistic Regression | Data Normalized | 78.5% |
Logistic Regression | Data Standardized | 80.5% |
2 . Decision Tree
Algorithms | Type | Accuracy |
---|---|---|
Decision Tree | Simple | 73.5% |
Decision Tree | Data Normalized | 43.0% |
Decision Tree | Data Standardized | 65.5% |
3 . SVM(Support Vector Machine)
Algorithms | Type | Accuracy |
---|---|---|
SVM | Simple | 86.5% |
SVM | Data Normalized | 62.5% |
SVM | Data Standardized | 68.5% |
4 . KNN(K Nearest Neighbour)
Algorithms | Type | Accuracy |
---|---|---|
KNN | Simple | 88.0% |
KNN | Data Normalized | 86.5% |
KNN | Data Standardized | 88.0% |
5 . Random Forest
Algorithms | Type | Accuracy |
---|---|---|
Random forest | Simple | 85.5% |
Randm Forest | Data Normalized | 73.0% |
Random Forest | Data Standardized | 84.0% |
From above results i conclude that KNN is best suited algorithm for this problem therefore finally i am applying KNN algorithm to get output