This assignment is focused on machine learning, mainly on the implementation of 2 different algorithms - Stochastic Grad

[answerhappygod]

This Assignment Is Focused On Machine Learning Mainly On The Implementation Of 2 Different Algorithms Stochastic Grad 1 (49.19 KiB) Viewed 43 times

This Assignment Is Focused On Machine Learning Mainly On The Implementation Of 2 Different Algorithms Stochastic Grad 2 (37.55 KiB) Viewed 43 times

This Assignment Is Focused On Machine Learning Mainly On The Implementation Of 2 Different Algorithms Stochastic Grad 3 (64.77 KiB) Viewed 43 times

This assignment is focused on machine learning, mainly on the implementation of 2 different algorithms - Stochastic Gradient Descent \& ID3 decision tree. The assignment is divided into two sections, each for one unique ML algorithm. The base structure and comments are provided on what should be done. You can use some libraries that help support you for the successful completion of the assignment. However, you CANNOT use a complete library that contains the implementation of ML algorithms. You can get pieces of code from online, but please cite the source properly.
Algorithm (60 points) Following are the data files that will be provided to you for the gradient descent algorithm implementation. - Training file - 'gd-train.dat' - Testing file - 'gd-test.dat' In these files, only non-space characters are relevant. The first line contains the attribute names. All the other lines are different example instances to be used for the algorithm. Each column holds values of the attributes, whereas the last column holds the class label for that instance. Write the code in the following code block, structure is provided. Instructions on the steps to follow are provided as comments.
Algorithm (90 points) Following are the data files that will be provided to you for the ID3 algorithm implementation. - Training file - 'id3-train.dat' - Testing file - 'id3-test.dat' In these files, only non-space characters are relevant. The first line contains the attribute names. All the other lines are example instances to be used for the algorithm. Each column holds values of the attributes, whereas the last column holds the class label for that instance. In a decision tree, if you reach a leaf node but still have examples that belong to different classes, then choose the most frequent class (among the instances at the leaf node). If you reach a leaf node in the decision tree and have no examples left or the examples are equally split among multiple classes, then choose the class that is most frequent in the entire training set. You do not need to implement pruning. Also, don't forget to use logarithm base 2 when computing entropy and set (0log0) to 0 . Write the code in the following code block, structure is provided. Instructions on the steps to follow are provided as comments. The code should output the following 3 things: