Month: April 2020

“Our company is using cutting edge machine learning technology.” Machine Learning vs. Data Analysis or Statistics From Dan Shewan on wordstream.com One big difference between data analysis and machine learning is the questions they seek to answer. In data analysis, you want to know something like what happened to sales at this point in time … Read more

It keeps your customers interested. In the earlier example, the music recommender was very accurate — it only recommended music from artistes you liked and previously listened to. A model solely focused on accuracy would do very well in offline evaluation (of the latest week of data). But in the real world, it would suck. … Read more

A template for configuring Flask + Gunicorn + Nginx + Docker with a detailed explanation, that should bring you a bit closer to working with microservices, building MVPs, and so on. Bo-Yi Wu via flickr It might be tricky to develop a good Machine Learning model, but even if one manages to do that, it’s … Read more

And why we might’ve gotten it wrong Source: Pexels About 10 months ago, a few of us began working on our model serving infrastructure. We wanted to build a tool that would take a trained model and turn it into a production web service, without us having to write glue code or wrangle AWS/Kubernetes for … Read more

MyReminderRepository.kt You can think of this class as a local server which means it serves us everything we needed to perform any action on the map such as adding a reminder or removing reminder and maintaining all reminders. The following code snippet creates an object of GeofencingClient which helps you to manipulate the geofences. private … Read more

Data preprocessing is a crucial step in machine learning and it is very important for the accuracy of the model. Data contains noise, missing values, it is incomplete and sometimes it is in an unusable format which cannot be directly used for machine learning models. But what if we use questionable and dirty data? What … Read more

Figuring out what we can do with the data available and what we can’t Photo by Alissa Eckert, MS, and Dan Higgins, MAMS, on PHIL Contents: Introduction Making our Chest X-ray COVID-19 classifier2.1. Data preparation2.2. Training2.3. Results Does it really work?3.1. Further analysis3.2. Discussion and takeaways Today everyone knows about the pandemic. Professionals do their … Read more

Null Elements: Since some users will not have values for certain features, they were captured as NULL. In actual sense, these null elements are zero values, considering aggregates were used. Hence, these null elements were replaced with zeros for all columns. Scaling: The range of values for all the features in the dataset, show quite … Read more

Notice how what we’ve gone through can be applied to learning almost anything. The most important takeaways being instead of focusing on what’s right at any given moment (impossible to predict), you’re concentrating on the trend. You’re building the habit of learning (using courses as a foundation) along with the habit of creating (building your … Read more

I truly believe that a Data and Analytics function has the mandate to enable better decision making within an organization. Very few practitioners would disagree with this argument, however for us to truly drive our vision, it is important to understand how people make decisions. The human decision-making process is ambiguous, to the extent of … Read more

Base rates, marginal probabilities, sensitivity, and specificity Photo by Markus Spiske on Unsplash When it comes to updating beliefs and making decisions under uncertainty, Bayes’ theorem is just about the best tool available. And yet it is so often relegated to academic textbooks and machine learning applications when it should be bringing us value in … Read more

Read in and split the Data For this analysis, I’ll be doing a random forest regression using the Boston Housing Dataset in the scikit-learn package. There are 13 features in the Boston Housing Dataset, you can read about them here. After we do some preliminary feature selection I’ll break down what the more important features … Read more

Or which forecast accuracy metrics to use? Source: https://www.arymalabs.com/ Many CPG brands across the world would be focusing on keeping a tab on their sales and demand numbers during the Covid-19 pandemic. In my previous article, I had covered points on doing Marketing Mix modeling during these testing times. The brands might have already forecasted … Read more

How to use Poisson regression and CatBoost to achieve better accuracy on count-based data… and predict the number of likes that a tweet gets. The concept of count-based data What is Poisson regression and why it is suitable for count-based data How to build a Poisson regression model with CatBoost package How to predict the … Read more

Deployment Now comes the crux of this post. This was the part of my journey where I couldn’t find much direct help on the internet. The Panel library (part of Holoviz) provides an excellent toolkit for managing data interactions, setting up pipelines, using widgets and deploying dynamic dashboards. Deployment is as easy as marking your … Read more

I am assuming that you have an AWS account, and you have access to the AWS Console. If not, you might need to sign up for an Amazon AWS account. First of all, we need to go to the Services tab to access the EC2 dashboard. 2. On the EC2 Dashboard, you can start by … Read more

The CORD-19 resource attempts to accelerate scientific discovery and save lives The network of diseases and chemicals associated with Chloroquine, an example of the kinds of insights that can be extracted from CORD-19 — this visualization was produced with the CoViz tool from AI2. The urgent phone call from the Michael Kratsios (whose august title … Read more

Listen to the data, curiously and carefully! Photo by Emma Frances Logan on Unsplash The fuel of each and every machine learning or deep learning model is data. Without data, the models are useless. Before building a model and train it, we should try to explore and understand the data at hand. By understanding, I … Read more

Is it possible for data science beginners to get up and running in under 2 minutes? Data science enthusiasts, how fast can you go from zero to Google Cloud Jupyter notebook? Let’s find out! Image: SOURCE. If you’re in the mood to ultra-customize your setup, Google Cloud gives you dizzying granularity. That’s a fabulous thing … Read more

A Generative model aims to learn and understand a dataset’s true distribution and create new data from it using unsupervised learning. These models (such as StyleGAN) have had mixed success as it is quite difficult to understand the complexities of certain probability distributions. In order to sidestep these roadblocks, The Adversarial Nets Framework was created … Read more

At the level of company-owned ANIs, an individual could best prepare themselves by being aware of how their data is collected (eg: cookies when you visit websites), used (eg: targeted advertising), stored (eg: Snapchat stores data on its servers), and biased at various levels (eg: biased training data, resulting in an inability for the service … Read more

What is the best strategy in order to find your perfect match? Is the first guy the best? Do we have to date as many guys as possible to find the best? But you can easily understand that these two strategies are risky and not optimal: in the first case, you can regret it later … Read more

SOLUTIONS FOR MICROSOFT POWER PLATFORM A step-by-step guide on the continuous delivery of AI Models, Power Apps, and Flows with Microsoft’s Power Platform and Azure DevOps. Photo by jesse ramirez on Unsplash In one of my recent stories, I’ve explained how to create a no-code AI prediction model using the Microsoft Power Platform to forecast … Read more

Hats off to Vicki for sparking the motivation to write this, I’m always open to ideas and suggestions for topics to write about *hinthint* If you came to me in 2010 and told me that email tracking analytics would be the hot stuff again in 2020, I would’ve thought you were crazy. But here I … Read more

As companies, particularly large organizations, scale up their models as a part of building an enterprise-wide pipeline, there’s an increasing need to operationalize the model development process. Similar to DevOps, models need to be developed, integrated, deployed and monitored. Often, with Enterprise AI initiatives, there are a host of governance considerations such as data integrity, … Read more

The NLP Project Basic Lexical Processing — preprocessing steps that are a must for textual data before doing any type of text analytics. Photo by Annie Spratt on Unsplash In this part of the series ‘The NLP Project’, we will understand the various preprocessing steps that must be applied before doing any type of text … Read more

Doing cool things with data! In this age of technology, data is the new oil. Organizations all over the world are transforming their environments, processes and infrastructures to become more data-driven. A major reason is that data analytics and machine learning gives organizations visibility into how to run their business better. The push to remote … Read more