Google Code-in

Linear Regression is the first step towards machine learning. Linear Regression is the most basic form of training your model, and then finding the output for a given number. You can learn about Linear Regression almost anywhere, there are various websites giving tutorials about them.

You have to make your own Linear Regression model using TensorFlow, first using a random data-set generated by python, and then later on the following sample data-set, which is a data from some company containing their salary v/s years of experience. You have to train a model on this data - set and see how much your model gives the output for someone with experience of 6.5 years.

When you complete this, you will get yourself familiarised with TensorFlow and its various methods. Don't forget to visualize your dataset, and then plot the regression line to see how much data falls on the sides of the line.