 In this learning dialogue, we will see the other two types of learning analytics that is predictive analytics and prescriptive analytics. Given that, you are performed diagnostic analytics such as why a student did not pass the exam or why the student was not able to answer a particular question. You found out that student did not pass the exam because he has not attended NF classes. Given that you have some diagnostic analytics like that. Can you think of how will you predict the student's performance in the future events? These diagnostic analytics you have performed based on the past data. Can you think of how do you perform in a future semester which student will pass the exam? Which student will not able to answer a particular question? You can pass this video, write down your activity, write down your answers. After you complete this activity, you can resume this video. Future analytics, actually predicting what will happen next. So, that is like what course will be performed better or which course will have less number of registrations in the university, which students will not complete the course. What will be the performance of student in a next question such a final level to the higher level like a course rate? How do you do that? Preparation is done based on the data from the past events. We collect the past events like how many students are registered to the particular course in the last 3 years. Can you predict it what will happen next class? Yes, if the students are registering to the course say networking for last 3 years only 60 percent students are registering then we can expect in next year also only 60 percent registers unless there is something event happens which makes the change. So, we can predict what will happen in a next semester based on the last past data. This is very common in a data analytics. We will talk about predictor analytics in a week 4 of this lecture. So, the most popular predictor analytics or NIE base or a very good class if it is called SVM or decision trees is very common in educational data mining also in the learning analytics community. There are lot of tools available for teachers and other stakeholders. The tools like VECA is open source for everyone and there are good tools which is developed for the industries such as rapid miner, orange. These software tools are available for the academic purpose free of cost. However, we have to understand one thing the latest bus about deep learning or artificial intelligence or deep neural network or conversion neural network. It is good to use those networks for the educational data, but we have to remember that in order to train those deep networks or CNN networks we need our lot of data. Usually in educational settings our data is fine-grained, but not lot of data on a similar event. So, we will stick to use NIE base SVM or decision tree. In our fourth lecture we will use one of these methods to predict the students outcome in a learning environment. The next activity is that is for the prescriptive analytics. Now that you are able to predict whether learner will continue their course or drop it right because you are able to do predictor analytics. Consider there are some learners who is going to drop your course. What measures will you take to ensure your learners are motivated enough to continue your course and how? Think about it, what measures will you take to ensure your learners to motivated enough to continue your course are not going to drop out. Prescriptive analytics can be scaffolding to help the students to achieve the learning goal. Scaffolding in the sense we can provide a hints feedback in the learning environment. For example in your class if you find couple of students needs help you can understand that these two students needs help and then they were not able to pass the exam based on the mid sem mark you can understand that. What you will do? You can ask them to do more assignments, you can give them more problems or you can do a special coaching class to them. So the students can clear the doubts by doing lot of assignments or you can combine them with the peers in the class they can interact and learn from others and they were able to complete the course. This classroom environment can be converted into intelligent learning environment using the data collected from students behavior and students interaction with the system. This is called personization or adaptation in the intelligent learning environment. You can predict the students current learning state and provide feedback or help to achieve the learning goal. Your learning goal can be said by you or the students. Students said their own learning goal you can help them to achieve their learning goal because student has to be motivated in order to complete the course.