Machine Learning with Python

The objective of the project is to gain some experience applying the Machine Learning techniques discussed in class. You will demonstrate your knowledge of these techniques in the context of a Supervised learning techniques such as Deep Neural Networks with Jupyter notebook, Theano, TensorFlow, Keras and Machine learning libraries in python on real datasets such as Novel Coronavirus or Titanic data. Data: Download the Novel Coronavirus dataset or titanic data set given below in the link, passenger ?patients survival data for the Titanic/Coronavirus [login to view URL] or [login to view URL] Perform the below Tasks using DNN or CNN or RNN I. Run a Deep Neural Network (feed forward neural network) with two hidden layers and ReLU activation function on the titanic dataset using the keras (10 Marks) a. Load the data from the csv file b. Build the Neural Network Classifier model using Keras c. Compile the model d. Setup Division of Data for Training, and Testing- 70% for Training and 30% for Testing. e. Train the Model f. Evaluate the Performance in terms of accuracy g. Make Prediction on Train and Test data h. Plot the accuracy results for training and validation(test) data i. Plot the loss for training and validation loss using Matplotlib j. shows the percentages of correct and incorrect classifications using confusion matrix. k. Plot the confusion Matrix. II. For the above Deep Learning Model, preprocess the data using Standard Scalar or MinMax Scalar, repeat the steps (b to k), Compare the Performance without preprocessing and with preprocessing and Tabulate your results and provide a justification which approach is better. III. Evaluate the Models using 10 folds Cross validation technique and compute the Accuracy of the model. Compare the results with train-test split and Cross validation, discuss which provides best performance IV. Prepare the Report discussing all the above tasks