Implemented and compared the performance of below Classifiers using Cross-validation & error metrics for classification Handwritten Digits.
- Linear classifier
- K-nearest neighbor classifier
- RBF neural network
- 1 & 2-hidden layer Neural Network
Notebook Descriptions:
-
Classification of Handwritten Digits using Logistic Regression Classifier.ipynb: Logistic Regression Modeled against scaled and unscaled data.
-
Change in Parameter Analysis for Logistic Regression.ipynb: Logistic Regression Performance analysis againt training set tuning & epoch tuning
-
Classification of Handwritten Digits using SVC.ipynb: SVC Modeled against scaled and unscaled data.
-
Change in Parameter Analysis for SVC.ipynb: SVC Performance analysis againt training set tuning & epoch tuning
-
Classification of Handwritten Digits using One Hidden Layer Neural Network.ipynb: 1 Hidden layer NN modelled and performance analysed for diff epochs, diff activation funcs & diff training set size
-
Classification of Handwritten Digits using Two Hidden Layer Neural Network.ipynb: 2 Hidden layer NN modelled and performance analysed for diff epochs, diff activation funcs & diff training set size
-
Classification of Handwritten Digits using Decision Tree .ipynb: Decision Tree Model of unscaled data.
-
Classification of Handwritten Digits using Decision Tree[max depth = 20] .ipynb: Decision Tree Model with max_depth = 20
-
Classification of Handwritten Digits using Decision Tree[max depth = 100] .ipynb: Decision Tree Model with max_depth = 100
-
Classification of Handwritten Digits using Decision Tree[train set = 30000].ipynb: Decision Tree Model with train_set = 30000
-
Classification of Handwritten Digits using Decision Tree[train set = 50000].ipynb: Decision Tree Model with train_set = 50000
-
Classification of Handwritten Digits using KNN Classifier [neighbors = 4 ].ipynb: KNN Model with 4 neighbors
-
Classification of Handwritten Digits using KNN Classifier [neighbors = 2 ].ipynb: KNN Model with 2 neighbors
-
Classification of Handwritten Digits using KNN Classifier [neighbors = 3 ].ipynb: KNN Model with 3 neighbors
-
Classification of Handwritten Digits using KNN Classifier [train data = 50000 ].ipynb: KNN Model with 4 neighbors and train_data is 50000
-
Classification of Handwritten Digits using KNN Classifier [train data = 30000 ].ipynb: KNN Model with 4 neighbors and train_data is 30000
-
Classification of Handwritten Digits using Random Forest Classifier [max_depth = 5].ipynb: Random Forest Model with 4 max_depth 5
-
Classification of Handwritten Digits using Random Forest Classifier [max_depth = 10].ipynb: Random Forest Model with 4 max_depth 10
-
Classification of Handwritten Digits using Random Forest Classifier[max_depth = 20].ipynb : Random Forest Model with 4 max_depth 20
-
Classification of Handwritten Digits using Random Forest Classifier[max_depth = 20][train = 30000].ipynb: Random Forest Model with 4 max_depth 20 and 30000 train data
-
Classification of Handwritten Digits using Random Forest Classifier[max_depth = 20][train = 50000].ipynb: Random Forest Model with 4 max_depth 20 and 50000 train data
-
Classification of Handwritten Digits using RBF Classifier{RBF}-[iter = 10].ipynb: RBF Model with epoch 10
-
Classification of Handwritten Digits using RBF Classifier{RBF}-[iter = 500].ipynb: RBF Model with epoch 500
-
Classification of Handwritten Digits using RBF Classifier{RBF}-[iter = 700].ipynb: RBF Model with epoch 700
-
Classification of Handwritten Digits using RBF Classifier{RBF}-[iter = 700][train = 30000]: RBF Model with epoch 10 and train data is 30000
-
Classification of Handwritten Digits using RBF Classifier{RBF}-[iter = 700][train = 50000]: RBF Model with epoch 10 and train data is 50000