opiaasg.blogg.se - Prodigy hack github get all the pets

You are not expected to be a code-monkey, but a part programmer, part researcher and problem solver. At advanced levels, you may be asked about Martingales and Markov chains.Ī very good place to get started with concepts are channels like 3blue1brown, sentdex and Siraj Raval on Youtube. You should not only be able to code but also be able to understand and explain the underlying principle, for example, how hypothesis testing works, significance of the p-value, and such. The focus areas are Python/R, Linear Algebra, Statistics (the more you can know the better). Even students with economics background make it big. Besides that, the best suited degrees are M. First of all, even if you are not from CSE background, it doesn't matter, but certainly helps to have such qualification. Value here refers to my view point and not what general population thinks. There is 0 value being a generalist in data science/Machine learning. My advice, get really good at math, stick to a toolset (R, Matlab, Julia). There are only very few organizations which are actually doing good in this area (google, facebook, baidu, microsoft). Then you can start to decide on approach.Ĭurrently the industry trend is to slap on Deep learning to everything and hope things work. Then you would know what to do to find the solution. Key thing for a data scientist is to understand followingįirst step in being a good scientist is to understand your problem domain. Some don't even write code by send across mathematical equations or excel sheets. They get thro' their job by doing R or Matlab (Yes, it is widely popular in industry). 90% of good data scientists have I have met don't understand programming or they don't even know what is Linux. (If you're not from CS I'll suggest not to go for those assignments)īut let me take a step back, most what is described in this post works great when you are not in part of industry but in college.

Andrew Ng's course on coursera and enroll. I think that's it guys! I'm open for questions.Īfter building the base go search for Prof.

Work out on your Database skills and learn how not to write spaghetti code.

Trust me this'll be the best thing you'll do to yourself if you're willing to come in this field! not sufficient! Search these MOOCs for individual courses on Regression/Clustering/NLP/etc. These will build the foundation which is obv. Ok so now I'll just post the suitable links to find the course! Also Probability, Stats and Number theory!īuy this or this or The Imposter Handbook or Grokking Algorithms (Introduction to Algorithms by Cormen is also a very good book!) Workout on your Maths! Learn Calculus and LA. CS50x by Harvard will be the best online free course for you.

So here's a Map that I made which "anyone" can follow to enter in this ML Industry. I'm making this post as because I've been asked this question many times!