Data Engineering and Mining

     Task description:   Data Engineering and Mining        The axioms set comes from the Kaggle Digit Recognizer rivalry. The aim is to allow digits 0 to 9 in handwriting images. Because the first axioms set is extensive, I keep frequently sampled 10% of the axioms by selecting the 10th, 20th examples and so on. You are going to use the sampled axioms to fabricate foretelling types using multiple tool acquirements algorithms that we keep scholarly recently: naïve Bayes, kNN and SVM algorithms. Tune their parameters to get the best type (measured by cantankerous validation) and assimilate which algorithms arrange emend type for this job.  Report structure:  Section 1: Introduction  Briefly portray the order total and unconcealed axioms preprocessing. Note that some axioms preprocessing steps perchance restricted to a point algorithm. Report those steps underneathneath each algorithm exception.  Section 3: Naïve Bayes  Build a naïve Bayes type. Tune the parameters, such as the discretization options, to assimilate results.  Section 3: K-Nearest Neighbor arrangement Exception 4: Support Vector Tool (SVM)  Section 4: Algorithm achievement comparison  Compare the results from the two algorithms. Which one reached loftier exactness? Which one runs faster? Can you illustrate why?