problem1. Apply [0.9, 0.7, 0.5] to the feed-forward neural network Figure 8.1 with associated connection weights in
a. Compute the input to nodes i and j.
b. Use the sigmoid function to compute the initial output of nodes i and j.
c. Use the output values computed in part b to determine the input and output values for node k.
d. Assuming the targeted output from K is 0.64, adjust all weights and using 0.5 for r.
You need to show computations of Error(K), Error(J), Error(I), delta Wik, Wjk, W1i, W2i, W3i, W1j, W2j, W3j, and all new weights.
problem2. And3.arff data set is attached. Use MultilayerPerceptron algorithm and use training set for test option to generate a model. If the accuracy is not 100%, adjust the parameters so that the model accuracy is 100%. Show your result in graphical format. Clearly indicate the weights of all edges.
problem3. Use CPU.arff dataset, Cross-validation test option, and Linear Regression algorithm. Show the last screen from Weka. What is your Linear Regression equation? Given MYCT=550, MMIN=8000, MMAX=22000, CACH=48, CHMIN=32, and CHMAX=64, what is the class value?
problem4. Use Contact-Lenses.arff dataset and Bayes algorithm. What is the probability of (young, myope, yes, normal) to wear none contact-lenses?