astion 3: We have a sentiment analysis problem of three classes i.e. Positive, Negative, and Neutral. The dataset contains a total number of 270 instances ( 90 of each class). Muhammad applied the machine learning classifier to this data. A train/test split ratio of 67 and 33 is used to carry out an evaluation of this data. Answer the question below: [2+2=4] 3.1: How can we represent training examples in these algorithms? 3.2: Which is the most common baseline evaluation measure to compare a proposed machine learning algorithm