Month: October 2020

Business Understanding and Data Understanding Let’s first take a look at the data. From the datasets, it is clear that we would need to combine all three datasets in order to perform any analysis. Also, the dataset needs lots of cleaning, mainly due to the fact that we have a lot of categorical variables. I … Read more

In simple terms, we can say that a vector is a quantity which has a magnitude as well as direction (elementary physics/math). In SVM; the magnitude is the numerical value (which can be calculated either by taking the mod or square root of sums of x and y coordinates; similar to hypotenuse in pythagoras theorum). … Read more

This post is a short summary and steps to implement the following paper: The objective of this paper is to learn self-supervised general-purpose audio representations using Discriminative Pre-Training. The authors train a 2D CNN EfficientNet-B0 to transform Mel-spectrograms into 1D-512 vectors. Those representations are then transferred to other tasks like Speaker Identification or Bird Song … Read more

Awesome Data is a GitHub repository with a seriously impressive list of datasets separated by category. It is updated regularly. Jeremy Singer-Vine’s Data Is Plural weekly newsletter has great fresh data sources. I’m always impressed by the quality. The archive is available here. In addition to competitions, Kaggle has a huge range of datasets. Kaggle … Read more

Introduction In traditional machine learning, to achieve state-of-the-art (SOTA) performance, we often train a series of ensemble models for combating the weakness of a single model. However, achieving SOTA performance often comes with large computation using big models with millions of parameters. SOTA models like VGG16/19, ResNet50 have 138+ million and 23+ million parameters respectively. … Read more

Required Packages The required python packages are as follows, (here I mentioned the packages with versions that I have used for the developments) Define Intents I will define few simple intents and bunch of messages that corresponds to those intents and also map some responses according to each intent category. I will create a JSON … Read more

To get the benefits of measure theory for the box volume the mathematician starts to work his way through the checklist. As a first step, he defines his box measure mathematically. A box in three dimensions can be written as a set: where w₁ ≤ w₂, h₁ ≤ h₂, d₁ ≤ d₂ are all real … Read more

Fairness in AI and AI bias Photo by Bill Oxford on Unsplash Artificial Intelligence has become an integral part of everyone’s lives. Starting from simple tasks like YouTube recommendations to complex life-saving tasks like generating drugs to cure illness, it has become omnipresent. It is influencing our lives in more ways than we realize. But, … Read more

An introduction to Pattern Exploiting Training Ever since the advent of transfer learning in Natural Language Processing, larger and larger models have been presented, in order to make more and more complex language tasks possible. But the more complex the models, the more time and the more amount of data it needs to train. The … Read more

Doing cool things with data The 2020 elections in the US are around the corner. Fake News published on social media is a HUGE problem around the election time. While some of the Fake News is produced purposefully for skewing election results or to make a quick buck through advertisement, false information can also be … Read more

Some of you may have noticed that it’s been a while since my last article. That’s because I’ve become a dad in the meantime, and I’ve had to take a momentary break from my projects to deal with some parental tasks that can’t (yet) be automated. Or, can they? While we’re probably still a few … Read more

In this article, I evaluate the many ways of weight initialization and current best practices. Initializing weights to zero DOES NOT WORK. Then Why have I mentioned it here? To understand the need for weight initialization, we need to understand why initializing weights to zero WON’T work. Fig 1. Simple Network. Image by the Author. … Read more

The picture above shows seven alternative elephant pictures and their DenseNet [9] classifications implemented in Keras with pre-trained ImageNet weights. As we can see, in many cases, the model recognizes the format (drawing, plush) instead of the animal in the image. Only two of the predictions has an elephant in the top5 predictions. The images … Read more

Prologue So, right now I want to test my skill using PyTorch. I am new to PyTorch, previously I am using Tensorflow to build an ML model. I already read all the basic tutorials and ready to build a real model using real data. What your favorite site to found real data to test your … Read more

We were excited to see the power of BigQuery receive a further boost this week with the release of 12 new BigQuery SQL features. Google Cloud describes these as “user-friendly SQL capabilities”. So, let’s take a look at what’s now possible. New to BigQuery is the ability to add new columns via the ALTER TABLE … Read more

Photo by Aung Soe Min on Unsplash And the difference between sort() and sorted() Dictionary is an important data structure that stores data by mapping keys with values. The default dictionaries in Python are unordered data structures. Like lists, we can use the sorted() function to sort the dictionary by keys. However, it will only … Read more

Tip #1: Communicate the problem as if it has nothing to do with Programming Polina Tankilevitch | Pexels Teaching basic algorithms can be one of the most challenging parts of introducing someone to Computer Science. Between the hypothetical problems and abstract thinking, it can feel overwhelming. But it doesn’t have to be. Unpopular opinion: Any … Read more

Pinhole Camera Model and Types of Distortion Before start correcting for distortion, let’s get some intuition as to how this distortion occurs. Here’s a simple model of a camera called the pinhole camera model. When the camera forms an image, it’s looking at the worldsimilar to how our eyes do. By focusing the light that’s … Read more

Step 7 : Add percentage of increase to answer reader’s questions We now are very tempted to quantify how much the 10% dropout accuracy helped over the 0% dropout setting. To help the reader for that, we add this percentage : Image by Author And we’re done ! The final matplotlib code is : Conclusion … Read more

Using PostgreSQL to create stateful, multi-page applications with Streamlit Photo by Mitchell Luo on Unsplash Streamlit Streamlit has come a long way since its inception back in October of 2019. It has empowered the software development community and has effectively democratized the way we develop and deploy apps to the cloud. However as with all … Read more

How to download files from the web, Bonus 2 In daily work life, we don’t always have the files we want to be saved locally. We might have to download them from remote servers or interact with them in place. While it is possible to download the files by hand, it would be a shame … Read more

According to the Pingouin homepage, this package is designed for users who want simple yet exhaustive stats functions. It was designed like that because some function, just like the t-test from the SciPy, returns only the T-value and the p-value when sometimes we want more explanation regarding the data. In the Pingouin package, the calculation … Read more

Now, we can move to the fun part, where we will write the edge detection function. You will be amazed at how simple it is using the OpenCV package. This OpenCV detection model is also known as the Canny edge detection model. Our function is consisting of three parts: edge detection, visualization, and lastly, saving … Read more

I recently adopted an extensive collection of notebooks that combined aid in the creation of analytics. It sounded like a daunting project to take on, but the more I read into the code, the more I realized it wasn’t all that bad. The notebooks looked overwhelming, but the code was relatively simple when broken down … Read more

Our Design — Based on separate building blocks working together For our system to scale and accommodate more features, more models and eventually more forecasters we divided it to the above building blocks. Each building block stands on its own but also knows how to communicate and work with others. A short explanation of each … Read more

A speedy intro to a Tableau feature that starts explaining the ‘Why’ of data points. No extra set up required. Image by chenspec from Pixabay One of the reasons data analysts create data visualizations is to find unexpected and unusual scenarios. The easier it is to get some more detailed information directly within the viz, … Read more

Before entering into details of how it is used for making a prediction, let’s describe the different elements that compose a Compact Prediction Tree (CPT): A trie, for efficient storage of sequences. An inverted index, for constant time retrieving of sequences containing a certain word. A lookup table, for retrieving the sequence from the sequence … Read more

Now that the project has been deployed to production, it’s a good time to look it retrospectively and reflect on various things/lessons that I learnt from doing it. 1. The training process There should be 3 main stages in predicting the contents of a form using the supervised learning approach, detect the text elements (printed … Read more

Developed the Algorithm Using a Real-World Dataset Logistic regression is a popular method since the last century. It establishes the relationship between a categorical variable and one or more independent variables. This relationship is used in machine learning to predict the outcome of a categorical variable. It is widely used in many different fields such … Read more