پروژه یادگیری ماشین با متلب

سلام علوم کامپیوتر... حدود دو هفته ماهکار سخت نیست برای متخصص ماتلابلطفا جزئیات کار بهش نشان دهیدWe want to compare these 3 types of learning (on the same data):1) isolated learning: just we run the same NN classifier on the data of each client separately (save the cross validation accuracy in text file, use more than one activation function, plot the results in bar chart).2) federated learning: the server trains the NN on his data, distributes the model to clients (only weight and bias, no shared data), each client starts his training from the obtained weight and bias (on his own data) and sends the new weight and bias to the server, the server aggregates all the weight and bias of all clients and train the NN on its data.... This is iteratively repeated until convergence. We aim to enhance the accuracy of the server's final NN model. The idea is by a collaborative learning without sharing the data, we enhance the final model at the server. Important: we need to save the accuracy, time, and confusion matrix of each training, to compare them later and prove the enhancement achieved by the federated learning.3) round robin learning: no central server but, the first client trains the NN on his own data and sends only the weight and bias to the next client. The second client starts from the obtained weight and clients and trains the NN on his own data, and so on... Also we need to save all accuracies to compare the learning types later (accuracy, time, confusion matrix).Finally we compare the three models... we aim the federated learning will be the bestNote: our data is in .xlsx file (row is observation, column is a feature, last column contains the class labels). The results are saved in text files, and .png bar chart plots.Allow me to specify 1) the input data file path, 2) the type of activation function and number of hidden layers (default is single), 3) select the learning type (isolated, federated, sequential round robin ). 4) the path of the output folder to save the text and plots..