Project

Part 1: What preprocessing steps were applied to the raw input data, and why were they necessary? Does your model exhibit signs of overfitting? If so, what steps would you recommend taking to address it? Is accuracy an appropriate metric for evaluating success in this case? Why or why not?

Part 2: What preprocessing steps were applied to the raw input data, and why were they necessary? How many distinct categories exist in this dataset? Does your model exhibit signs of overfitting? If so, what steps would you recommend taking to address it? Is accuracy an appropriate metric for evaluating success in this case? Why or why not?

Part 3: What preprocessing steps were applied to the raw input data, and why were they necessary? Why is this problem categorized as regression rather than classification? Does your model exhibit signs of overfitting? If so, what steps would you recommend taking to address it? Is mean absolute error (MAE) a suitable metric for evaluating success in this case? Why or why not?

TODO 1 Refer to "Deep Learning with Python" by François Chollet for an introductory guide to deep learning and its practical applications, including sentiment analysis and text classification. Explore the provided datasets for binary classification of movie reviews and multiclass classification of newswires, as well as regression for estimating house prices.

TODO 3 Implement code to showcase two cases: one false positive and one false negative made by the classifier. Display both the true value and the predicted value. Comment on the errors and potential causes for each misclassification.

TODO 4 Experiment with different configurations of:

• Number of hidden layers and units per layer
• Loss functions
• Activation functions Avoid using complex architectures and advanced techniques to mitigate overfitting.

TODO 5 Experiment with different configurations of:

• Number of hidden layers and units per layer Avoid using sophisticated architectures and advanced techniques to prevent overfitting.

(OPTIONAL) TODO 6 Explore various configurations of:

• K for K-fold validation
• Number of epochs
• Number of units per hidden layer
• Number of hidden layers Avoid using complex architectures and advanced techniques to mitigate overfitting.

Import necessary libraries such as matplotlib.pyplot for visualization and keras for building neural networks with specified layer configurations.