Is it hard to find Civil Engineer jobs?

What should I do to find jobs in Machine Learning, if I am Software Engineer at Indian Software Companies?

• I am currently working at Software Industry. I have 3 years of experience in software development. Modules which i have developed were mainly modules for web application for banking project. I am not very happy with my job.I am interested in learning Machine Learning Algorithms, Data mining. What should I do to change my job track. Is it even possible to change track ?

• Answer:

  Ahh.. change track.. ofcourse you are responsible for you life and what you love. If you are going for good pay which ML jobs get then this is not a good decision or maybe yes.. you are hardworking enuf to get there. Its totally upon you what you want and are you ready to do what it takes.   I myself have worked only for 3 months in a product development company and dont have an ML job. So I may be inexperienced to say anything. But I know decent ML. But what I learn from Internet is. Many of the ML jobs use basic ML stuff and a lot of data cleaning and getting data ready for ML stuff. According to me it is not the part of exciting ML field(however still important). Real exciting ML jobs does constitute so called interesting ML stuff and allow for daring new things untested however it may be resource consuming including time. Maybe u need to rewrite the code in order to parallelize it for large realtime data flow(maybe from Internet). good ML skills = better CS skills than a statistician + better statistics skills than a Computer scientists. Finally on skills needed. You should know some of no-sql database, Hadoop, data visualization skills, a low level language like C(with matrix libraries like BLAS and its decendents), a high level language like python(with matrix libraries like numpy and scipy) or MATLAB(which is made for matrix manupulations). Machine learning theory and its implementation details. sometimes understanding of Mapreduce is also expected. If you want to be bayesian(one of the two broad classes Frequentist and bayesian statistics) and want to do inference then some approximate inference tools. Apart from this ML itself is very vast and need to be understood well. Few things you should know is Linear algebra, a mixture of bayesian and frequentist methods, optimization, learning theory is generally not needed but is good to know. This is my favourite picture of ML books. lot of ML resources here. (http://dk-techlogic.blogspot.in/2012/05/best-machine-learning-resources.html) All though duda hart is a classic text do not start with that. Machine Learning a Probabilistic Perspective by keving murphy is my favourite. I sounds a lot and it is. You really need to love ML to do all of these. You are not happy with you job is not enough. You need much more motivation than that. You need to be a learner , an explorer and open to non-conventional ideas. My personal perspective: I was in a product comapny and wanted to learn ML. I left my job right over to learn ML. Went to an academic lab as an intern where I could learn ML and I know people who have been in software industry for 6 years doing the same. As of from their perspective, they have the technical skills and if they know the theory by devoting some time, they are good to go.. Hope this helps.... :)