Monday, April 28, 2014

May Goals: Math!

Not even on the 2nd part of the Andrew Ng's Machine Learing course do I feel my lack of math expertise hindering my understanding =(.

Background

I really slacked off during my middle school / high school days. Prime example of a procrastinator. Failed AP Calc, passed Calc and Stats in college, and never took another math class again. I really did enjoy stats, much more practical in day to day life.

There are gaping holes in my math abilities. I usually did well enough to pass, but that's horrible since this left me with a swiss cheese foundation. Probably didn't retain 90%+ of what I "learned" in school. I think the biggest problem is that there was no connection to how applicable this is to the "real world", most likely why I liked stats, I saw that it was useful in determining bias with data that we're constantly bombarded with.

May Goal 

Utilizing Khan Academy I plan to get through Algebra II, maybe even Probability and Statistics, by the end of May. Courses I'm going to skip include Geometry and Trig since I consider them vastly difference than what I think is relavent for Machine Learning, unless I'm going to concentrate on Computer Vision. I have to consider whether or not to do Calculus, this will largely depend on what the Prob/Stats course will cover.

Did I mention that I love MOOCs? Well I do! <3.

Saturday, April 26, 2014

Google's Making Sense Of Data = Fusion Table Tutorial?

Finished all 3 units of Google's Making Sense of Data course today. It took about a week to finish without doing the final projects.

What I was expecting: Introduction to data science and learning statistical methods.

What I got: Introduction on how to read graphs, empirical research, and Fusion Table.

This was a super basic course. I wouldn't recommend to anyone that has taken basic stats already. This course really acts as a tutorial for Fusion Tables, a Google product that's an extension of their Spreadsheet. Fusion Table feels clunky, but the I love how easy it was to visualize the data, I mean, we even get heat maps. =D


I like how they UI of the course probably better than the Edx course, but the layout is practically identical.

Gonna see if Andrew Ng's ML class has a stats prereq so I can start diving in and create that Singularity.

Saturday, April 19, 2014

April Goals Completed (...Somewhat lol) / Testing Careers

MITx

One the big things I wanted to pick up from the MITx course was how to determine O notation (time complexity). The course did a fairly good job of explaining it and I passed most of the quizzes on it, though my n log n understanding still needs work.

The last weeks of the course covered decision trees. I'm going to agree with commenters that this section feels rushed and was quite confusing for me to understand what it is suppose to do to writing code for it.

I did skip a few things like classes and object oriented programming since I did most of it during the bootcamp I attended.

Overall: This is a solid overview of python and how to develop time efficient algorithms. I think this covers a lot more detail than just the Codeacademy course on Python.

Testing Careers

I think going through the bootcamp was a great way to figure out if I wanted to do software development as a career, to a lesser extent I think this course helps with that decision as well. For me, these types of courses helps connect the dots that traditional school wasn't able to do. I learned skills from the courses, then was made to complete assignments that can be applied to real world situations right away.

The next courses I'm looking to take are in Machine Learning and Data Science. Since most investments/decisions are data driven these days, having knowledge of stats is super important. I enjoyed statistics when in school because it seemed so practical, especially in reading studies that we're bombarded with everyday and knowing what the conclusions actually mean and questioning methods that a study uses.

The bright side for me is, a lot of Machine Learning and Data Science in companies require programming skills as well. To develop the tools to analyze information, storing and pulling data from databases, and parsing documents.