Ahmed T. Hammad
ā¹ļø About
š§āš« Teaching
š° Research
š§āš Students
ā Papers
š”Solutions
š§°Toolkits
š CV
šļøBlog
š½ļøļø Slides
Github
LinkedIn
Email
šļøBlog
A Fresh Perspective on Treatment Effects - Beyond the Average and Into the Tails
Most evaluation of interventions ā policies, programs, experiments ā centers on the Average Treatment Effect (ATE). Did the treated group do better on average than theā¦
All About MLflow
If youāve spent any time doing machine learning seriously, youāve run into this problem: you trained a model last week that performed better than anything you have now, andā¦
Contextual Multi-Armed Bandit: Maximizing Rewards with Intelligent Decision-Making
Picture a row of slot machines. Each has its own payout probability, and you donāt know any of them upfront. Your goal is simple: walk away with as much money as possible.ā¦
Data Science Books
Books I actually recommend to people, with honest takes on what each one is good for.
Data Science project Boilerplate
Every time I start a new data science project, I go through the same setup steps. Create folders, set up the virtual environment, add a .gitignore, write the Dockerfile. Itā¦
Developing in a Docker container
I develop inside Docker containers. Not because itās trendy, but because it solves a real problem: my local machine stays clean, the project environment is reproducible, andā¦
Embracing Change: Incremental vs. Batch Machine Learning
Most machine learning tutorials train a model, evaluate it, and stop. Thatās fine for a homework assignment. In production, itās usually not how things work. Real systemsā¦
From ATE to Uplift Modeling
In a randomised controlled trial (RCT) the standard output is the
Average Treatment Effect (ATE)
: one number telling you how much the treatment moved the outcome on average.ā¦
How I Created R-Genius: A Journey into Empowering R Users with AI
R is one of the most powerful tools in data science. Itās also one of the most unforgiving. Error messages that reveal nothing, package ecosystems that overlap in confusingā¦
Logistic Regression and Marginal Effects
Logistic regression is everywhere in applied data science ā binary outcomes, classification problems, probability estimation. Most people know how to fit one. Fewer know howā¦
Machine Learning, Copula and Synthetic Data
Synthetic data is one of those ideas that sounds like it shouldnāt work ā if the data is fake, how can a model trained on it generalize to real data? The answer is thatā¦
On Learning Methods in the Age of AI
In recent years, something subtle has changed in how students approach programming. What used to begin with a blank script and a vague idea now often begins with a prompt. Aā¦
Online Uplift Modelling with River
The usual workflow is:
Probability Box with Kernel Density Estimation
Weather forecasts got me thinking about data. A simple historical table ā temperature, humidity, rain ā and the question: given all this data, what can I actually say aboutā¦
Quantile Random Forest
Most regression models give you a single number: the expected value of the target given the inputs. Thatās often what you want, but it throws away information. Quantileā¦
The 3 + 1 pillars of data science
A few weeks ago, one of my students posed a question I wasnāt expecting:
The Beauty of Soft Decision Trees
Decision trees work by making hard choices at each node: if
\(X_1 > 5\)
, go left; otherwise, go right. Itās clean, interpretable, and brittle. A point sitting right on theā¦
The R
rgcapi
library
Algorithmic trading has always interested me. The abundance of time-series data, the clear feedback loop, the challenge of building and testing strategies ā itās a domainā¦
Understanding Stationary: Concepts, Implications, and Approaches
Time series analysis runs through economics, finance, engineering, and the natural sciences ā any domain where observations are indexed in time and the ordering matters.ā¦
Unraveling the Power of Causal Machine Learning
Standard machine learning is very good at finding patterns. Given enough data, a model can identify that A and B tend to co-occur, that X is predictive of Y, that a certainā¦
Updating knowledge with Bayes
Iāve explained Bayesian updating many times over the years ā to students, to colleagues, to people with no statistics background at all. The examples I find work best areā¦
When in doubt, just model it. Modelling uncertainty
Thereās a pattern Iāve noticed across projects: when something is hard to measure, people tend to either ignore it or collapse it into a single number. Both choices feelā¦
pbox: Exploring Multivariate Spaces with Probability Boxes
In a previous post I introduced the idea of a āprobability boxā ā turning a dataset into a queryable probability space using Kernel Density Estimation. That was theā¦
No matching items
