A generic framework to segment customers more naturally

Image for post
Image for post

Intro

Businesses have this ever-lasting urge to understand their customers and it kinda makes sense. The better you understand the customer, the better you serve them, and the higher the financial gain you receive from that customer. This is not a new topic of discussion in the industry. Even since the dawn of trade, this process of understanding customers for a strategic gain has been there practiced and this task is known majorly as “Customer Segmentation”.

Well as the name suggests, Customer Segmentation is a task-agnostic term that covers all the ways, a business could segment its customers according to their precise needs. Some of the common ways of segmenting customers are based on their Recency-Frequency-Monatory values, their demographics like gender, region, country, etc, and some of their business-crafted scores. …


Testing is an essential component of effective outbreak responses. Without widespread testing, we cannot know whether a disease is spreading nor take measures to appropriately respond to it. “All countries should be able to test all suspected cases, they cannot fight this pandemic blindfolded, they should know where the cases are, and that is how they can take decisions,” said Dr. Tedros, WHO Director-General.

But to ramp up testing to a much-needed scale, you need the laboratory space as well as enough, and the right kind of, machines. You need the right reagents — highly specific substances used to extract the virus’s genetic material and to make it easier to study. You need staff to take the swabs from patients’ noses or throats, and staff in labs to process the tests. And you need the logistics in place to get samples from patients to labs. We’re talking about diagnostic tests to find out if you have the virus here — ones that involve a nose or throat swab that has to be sent off to a lab. All this could be difficult to mobilize for some developing nations. India has increased its coronavirus testing capacity to 10,000 samples per day. …


Image for post
Image for post

Hyper-parameter tuning in machine learning models at scale using Pyspark for free

After months of chasing and scrutiny, finally, you got access to the data dump you have been waiting for. You are already done with the pipeline of necessary preprocessing steps, some feature engineering and some model selection. After a bit of experimenting, you picked your best algorithm to proceed with, execute the hyperparameter tuning part, fill up your big cup of coffee and you wait, and wait, and wait…….and wait for the results like a never-ending story. This is a vastly common phenomenon in any typical Data scientist/Machine learning engineer’s daily routine. I am sure it’s considered the most boring and monotonous task in the whole ML life cycle if there is any. I stumbled upon a cool way that could reduce this waiting time to a great extent. It’s pure magic. It’s like after reading this “You are a wizard Harry!”. Anyway, the magic’s got a name and it’s…..wait for it….. “Parallelization”. …


Image for post
Image for post

In this post, I will explain how to use the vanilla version of YOLOv3 to detect objects from the COCO dataset and also how to custom train your own dataset for your own use-case.

The Yolo detection code here is based on Erik Lindernoren’s implementation of Joseph Redmon and Ali Farhadi’s paper.

Here are the links for the series.

All about YOLOs — Part1 — a little bit of History

All about YOLOs — Part2 — The First YOLO

All about YOLOs — Part3 — The Better, Faster and Stronger YOLOv2

All about YOLOs — Part4 — YOLOv3, an Incremental…


Image for post
Image for post

YOLOv2 project started to make YOLO better in terms of accuracy, faster in terms of speed and stronger in being able to classify more classes.

This 5-part series aims to explain everything that is there about YOLO, it’s history, how it’s versioned, it’s architecture, it’s benchmarking, it’s code and how to make it work for custom objects.

Here are the links for the series.

All about YOLOs — Part1 — a little bit of History

All about YOLOs — Part2 — The First YOLO

All about YOLOs — Part3 — The Better, Faster and Stronger YOLOv2

All about YOLOs — Part4 — YOLOv3, an Incremental…


Image for post
Image for post

Before YOLO there were two major object detection frameworks, DPM(Deformable parts model) and R-CNN both region-based classifiers where, as a first step they would find regions and for the second step, pass those regions to a more powerful classifier to get them classified. This approach involved looking at images thousands of times to perform detection. YOLO started as a project to optimize this approach by building a single neural network that takes a single image and gives back the detections and class in a single pass. That’s why the pun “You Only Look Once.”

This 5-part series aims to explain everything that is there about YOLO, it’s history, how it’s versioned, it’s architecture, it’s benchmarking, it’s code and how to make it work for custom objects. …


Image for post
Image for post

I am pretty sure; every computer vision enthusiast must have heard about YOLO, a real-time object detector that can identify the objects and where those objects are in an image or a video. Apparently people talk about either one version of YOLO or partial concepts of YOLO or how to make it work with code. I couldn’t find a single source on the internet that talks about YOLO in its entirety. In comparison to recognition algorithms, a detection algorithm, in this case, Yolo, does not only predict class labels but detects locations of objects as well. …


Image for post
Image for post
Pic credits: lynda.com

Deep learning-based Anomaly Detection using Autoencoder Neural Networks

In generic terms, anomaly detection intends to help distinguish events that are pretty rare and/or are deviating from the norm. This is of high importance to the finance industry like in consumer banking, anomalies might be critical things — like credit card fraud. In other cases, an anomaly might be something that companies look for to leverage from it. Some of the other applications include Intrusions in communication networks, Fake news, and misinformation, Healthcare analysis, Industry damage detection, Manufacturing, Security and surveillance, etc.

The use-case shown in this article is from the SAP domain particularly, Finance. …


Image for post
Image for post
Credits: https://roosboard.com

a primer for old school businesses to venture into AI

Artificial Intelligence is a buzz word you might have probably heard of a lot, lately. You being an expert in your business world might be thinking…

“is it something I should be experimenting with?”

“is it going to put me out of business?”

“is it significant enough change for me to worry about?”

“is it something which if I early-adopt is going to give my business a competitive edge?”

Well, this article doesn’t answer these questions but it will give you an intuitive understanding about AI without getting technical which will indeed make you answer them yourself. …

About

Rehan Ahmad

Senior AI Expert at Qualcomm

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store