Data Wrangling

Data cleaning. Also known as data munging. A process that transforms raw and problematic data (eg. missing values) into a form that can be used for analysis or models downstream.  

GPU Computing

GPU computing uses GPU as a co-processor to accelerate CPUs for general-purpose engineering and scientific computing. It accelerates applications running on the CPU by offloading compute-intensive and time-consuming segments of the code. This is also called hybrid or heterogeneous computing.

Neural Network

Neural networks consist of algorithms modeled on the human brain. These algorithms use machine perception to calculate and interpret sensory data. These networks also identify, label, and raw cluster data, based on similarities among them, in labeled training datasets. This is just like our brain does for decision-making.


NumPy is a Python library that adds support for huge, multidimensional arrays and matrices, and high-level mathematical functions that operate on these arrays.

As Python is interpreted, NumPy in Python provides functionality similar to MATLAB, and they both allow users to construct fast programs as long as most operations are performed on arrays or matrices rather than scalars.


A python library used by data scientists, engineers, and non-data scientists that manipulates tabular data. Think of it as excel on steroids. Pandas works with hundreds of thousands of rows without performance degradation and users can query data to answer questions.      

Parallel Computing

In parallel computing, larger problems are broken down into smaller, independent parts. These smaller parts are executed simultaneously by multiple processors that communicate via shared memory. The results of these computations are then combined to form the result of the original larger computation. It speeds up application processing and problem-solving.


The building block for linear algebra. The meaning changes in the fields of physics, mathematics, and computer science. In physics, it’s defined by the length and direction a line is pointing to, which can move around anywhere in space and still be the same vector. In computer science, it is an ordered list of numbers. In math, it “generalizes” the two, a “vector can by anything where there is a sensible notion of adding two vectors and multiplying a vector by a number.”      

Word Embedding (Word Vector)

The process of applying a “numeric vector input that represents a word.” For example, the letters of the word “the” can be represented by the vector length 500. And every word used in the vocabulary would have a unique vector. This is a method used in NLP (natural language processing).

Scroll to Top