Everything that you have mentioned is good. I would really focus on understanding a wide range of traditional statistics, ranging from various descriptive and

Everything that you have mentioned is good. I would really focus on understanding a wide range of traditional statistics, ranging from various descriptive and exploratory techniques, to dimension reduction and latent factor discovery as well as a variety of regression techniques. Then dig into the whole machine learning space as you already suggested. I would also get very comfortable with SQL - much of the data that you will work with in an enterprise setting will be accessible through SQL querying or some close variant. If you come from a programming background Python will be more sensible and accessible for you as you learn. If you come from a more classical research background then R will generally be a more comfortable place to start. Enjoy the journey