Two datasets are provided.
(a) The training set (training.data): Each line in training data is a sample containing a class label and the features that are separated by a comma. For
(b) The test set (test.data): Each line is a sample without its class label.
Task: Use the training set to build a classifier (limited to what we have covered in our class), and apply it to the test set and submit your prediction
Copyright By PowCoder代写 加微信 powcoder
results in a text file (one prediction per line). The order of your predictions should be the same to the order of test samples. In addition, you should
also submit a description of what you do to get your results, which should be detailed enough for readers to reproduce your results.
Grading: Report 50 points. We will assume that the prediction accuracy follows a Gaussian-like distribution. The mean and variance of all predictions
will be first calculated and be used to calculate your accuracy scores. Those higher than (mean + 2 std) will get 50 points, those equal or lower than
(mean – 3std) or random guess will get 25 points, and the rest in between will get a score based on the scale decided by the first two settings
程序代写 CS代考 加微信: powcoder QQ: 1823890830 Email: powcoder@163.com