Brief explanation on CNN

The following brief article is part of a lecture on CNNs at Stanford University.

A very good explanation of the technique.

 

Posted in Artificial Intelligence, Uncategorized

The Dark Secret at the Heart of AI

Article in MIT Technology review explains the need for the explanability of a deep learning system for the decisions it makes. This is important in various critical domains such as healthcare and military. Unless the machine does not explain its reasoning better than humans, do not trust it.

The Dark Secret at the Heart of AI

Posted in Articles, Artificial Intelligence

Python Recipes

Recently started to take Python seriously.

I was looking for a recipes section similar to C++ . I found this to be a good resource

Hettinger’s recipes on Python

 

Posted in Programming

Datasets

These datasets have been curated by Data Scientists for their research and are available for general use.

Follow link to retrieve the datasets.

http://blog.mortardata.com/post/67652898761/6-dataset-lists-curated-by-data-scientists#!

Posted in Datasets

TED Talks on Visualization

Here are some TED talks about the field of visualization in general.

http://www.ted.com/playlists/56/making_sense_of_too_much_data.html

Posted in Uncategorized

An introduction on the blog

Finally got around to writing my first post on the blog. The last quarter of 2012 was one of the most eventful months of my life. I got a well-paying job, finished my PhD, got married, and moved to a better location . The blog took a back seat and also my reasons for writing this blog.

One of my main motivations for writing this blog is professional. This is an age of the digital self. You are online more than you realize. I wanted to keep an updated blog to talk about and share information about my professional interests . Scientific visualization was my chosen field of research the past several years. However, I found myself over the last year exploring the parallel field of data visualization and analytics and fell in love with it. The creative techniques used to display data and information was impressive and over the past year I was completely engrossed reading the background. I used these new concepts within my field of research in medical image analysis to gain quite interesting information from the data that I never thought existed. Someday I will post my results but not here at this time.

The data of data visualization has evolved into an exploratory medium where collective intelligence can be gained through sophisticated  data manipulation techniques. These techniques target the perception of the human mind and indirectly help our cognition. Until very recently, these techniques have come into the mainstream through the introduction of big data . The stream is multidisciplinary and will require more exposure to databases, statistics, machine learning, and mining. Also required is a good amount of engineering skill to derive new algorithms for effective data analysis.

All these different fields are vast and requires a good amount of learning time, and I was not sure how would I be able to get an introduction to these topics in a way that will allow me to explore further. This changed when I decided to take on the online course by Alberto Cairo (which I recommend to anyone who is interested in data visualization). I had a good experience taking that course and decided to explore further with courses from Coursera and EdX. I am very impressed by MOOCs in general. It allows one to explore topics that would have taken substantial time to research on your own. My main motivation was to gauge my interest on the subject material. The only cost that one accrues is the time you put in. I think the medium of online education should be a great motivational factor for undergrads deciding on a career planning mission.

As I build my skills towards this field, my goal with this blog is to explore datasets and visualize, analyze, and summarize the results. The reason for doing this is to document the research for myself and others in the same path as I am. Moreover, it will be a good reference material for the future. This will also be a good exercise to use new technologies outside of the general languages that I have been using over the years.

Finally, to end my first post I hope this endeavor will be worth it and I will be better at what I am trying to achieve. Until then I am enjoying my journey into the Wild Wild West of data visualization and analysis.

Posted in Uncategorized