Essentials of Deep Learning: Introduction to Unsupervised Deep Learning (with Python codes)

In one of the early projects, I was working with the Marketing Department of a bank. The Marketing Director called me for a meeting. The subject said – “Data Science Project”. I was excited, completely charged and raring to go. I was hoping to get a specific problem, where I could apply my data science wizardry and benefit my customer. The meeting started on time. The Director said “Please use all the data we have about our customers and tell us the insights about our customers, which we don’t know. We really want to use data science to improve our business.” I was left thinking “What insights do I present to the business?” Data scientists use a variety of machine learning algorithms to extract actionable insights from the data they’re provided. The majority of them are supervised learning problems, because you already know what you are required to predict. The data you are given comes with a lot of details to help you reach your end goal. Source: danielmiessler On the other hand, unsupervised learning is a complex challenge. But it’s advantages are numerous. It has the potential to unlock previously unsolvable problems and has gained a lot of traction in the machine learning and deep learning community. I am planning to write a series of articles focused on Unsupervised Deep Learning applications. This article specifically aims to give you an intuitive introduction to what the topic entails, along with an application of a real life problem. In the next few articles, I will focus more on the internal workings of the techniques involved in deep learning.

Six Sigma DMAIC Series in R – Part 2

Hope you liked the Part 1 of this series. In this part, we will go through tools used in Measure phase of DMAIC cycle. In this phase, the baseline and target performance of the process must be determined and the measurement systems are validated. Two most representative tools during measure phase are: (1) Measurement system analysis (2) Process capability analysis

Machine Learning Explained: Vectorization and matrix operations

Today in Machine Learning Explained, we will tackle a central (yet under-looked) aspect of Machine Learning: vectorization. Let’s say you want to compute the sum of the values of an array. The naive way to do so is to loop over the elements and to sequentially sum them. This naive way is slow and tends to get even slower with large amounts of data and large data structures. With vectorization these operations can be seen as matrix operations which are often more efficient than standard loops. Vectorized versions of algorithm are several orders of magnitudes faster and are easier to understand from a mathematical perspective.

What is a Tensor?

A tensor is a container which can house data in N dimensions, along with its linear operations, though there is nuance in what tensors technically are and what we refer to as tensors in practice.

WTT: What the tensor?

Ron Bodkin explains what a tensor is and why you should care.

Remove password protection from Excel sheets using R

Most data scientists wished that all data lived neatly managed in some DB. However, in reality, Excel files are ubiquitous and often a common way to disseminate results or data within many companies. Every now and then I found myself in the situation where I wanted to protect Excel sheets against users accidentally changing them. A few months later, however, I found that I sometimes had forgotten the password I used. The “good” thing is that protecting Excel sheets by password is far from safe and access can be recovered quite easily. The following works for .xlsx files only (tested with Excel 2016 files), not the older .xls files.

Send tweets from R: A very short walkthrough

There are a few reasons why you might want to send tweets from R. You might want to write a Twitter bot or – as in my case – you want to send yourself a tweet when a very long computation finishes.

Deep Learning from first principles in Python, R and Octave – Part 8

This post, Deep Learning from first Principles in Python, R and Octave-Part8, is my final post in my Deep Learning from first principles series. In this post, I discuss and implement a key functionality needed while building Deep Learning networks viz. ‘Gradient Checking’. Gradient Checking is an important method to check the correctness of your implementation, specifically the forward propagation and the backward propagation cycles of an implementation. In addition I also discuss some tips for tuning hyper-parameters of a Deep Learning network based on my experience.