XGBoost Explained: From Gradient Boosting to Weighted Quantile Sketch
XGBoost Introduction: XGBoost XGBoost is a tree-based algorithm. It is a variant of Gradient Boosting which is one of the highest-scaled versions of…
Latest News
⚑ Report a ProblemTech news from the best sources
All topics
AI
Gear
News
Tech
agents
ai
api
architecture
automation
beginners
career
database
devchallenge
devops
javascript
llm
machinelearning
mcp
opensource
performance
productivity
programming
python
react
security
showdev
testing
tutorial
typescript
webdev
Where Did This Trade Come From? Attributing DEX Volume in Dune SQL
This is an adapted excerpt from Chapter 6 of my book, The DeFi Data Quick-Start: Reading On-Chain Truth with Dune Analytics . The book goes from you…
Backtesting overfitting: why your backtest lies and how to make it honest
Cross-post. Original: stellarbytecapital.com/blog/backtesting-overfitting A profitable backtest is the easiest thing to produce in all of quant trad…
Macaron-V1: Continual Learning with Self-Improvement and Mixture-of-LoRA Adapters
This is a Plain English Papers summary of a research paper called Macaron-V1: Continual Learning with Self-Improvement and Mixture-of-LoRA Adapters…
BDH-CQ Uses Recurrent Latent Reasoning to Cut ARC-AGI Inference Costs
This is a Plain English Papers summary of a research paper called BDH-CQ Uses Recurrent Latent Reasoning to Cut ARC-AGI Inference Costs . If you lik…
How Cross-Model Compatibility Lets Attackers Extract Proprietary LLM Reasoning Traces
This is a Plain English Papers summary of a research paper called How Cross-Model Compatibility Lets Attackers Extract Proprietary LLM Reasoning Tra…
Beyond the Basics: What I learned from a Deep-Dive EDA on the Ames Housing Dataset
Hey everyone! As I’m working through my ML fundamentals, I wanted to share a deep-dive Exploratory Data Analysis (EDA) I just completed on the class…
How late is Form 4 'real-time' data? Measuring filing delay on 11,241 insider filings
If you backtest anything involving insider trades, there's a quiet bug waiting for you: Form 4 data is timestamped twice . Every filing carries a tr…
Building an Open Turkish EV Charging Intent Dataset
Electric-vehicle assistants rarely have just one job. A short Turkish question may ask for a nearby station, a charging-price comparison, help plann…
Same Zone, Same Projection, 427 Metres Apart
I was writing a blog post about the architecture of a system I built when I found a bug in it. This is that bug, and the much more interesting thing…
Why Your ML Model Should Sometimes Say 'I Don't Know': Building Uncertainty-Aware Clinical AI
🚨 Predictive models tell us what will happen. 🧠 Causal models tell us what happens IF we intervene . ❓ But neither tells us what we should DO . In h…
Understanding the Git Workflow: Working Directory, Staging, Commit and Push.
Two weeks ago, I joined a Data Science bootcamp and part of the learning outcomes for this week was independently navigating the Git Workflow to suc…
Why Statistics Is the Backbone of Data Science
Statistics isn't a supporting subject in data science, it's the thing that makes the "science" part of data science actually mean something. Code Gi…
Correlation Is Pairwise. Multicollinearity Isn't.
Almost every regression project I've seen in the last decade starts the same way. Load the data, df.corr() , plot the heatmap, scan for red squares,…
Why I Stopped Rushing to Machine Learning and Started with SQL Instead!
I've solved 15 of the 50 problems in LeetCode's SQL 50. I started my first end to end ML project predicting telco customer churn this week. And thre…
Tracking Football Odds Movement in Python — the Full Tick History, Not Just a Snapshot
Most football data APIs will tell you what the odds are right now . A few keep yesterday around. Almost none will show you the full journey — every…
I built a crypto trading bot. It lost to doing nothing.
Every trading bot post ends the same way: a green equity curve and a suspicious silence about whether it made real money. This post is the other one…
Beyond the Vector: Why Graph Neural Networks are the Strategic Choice for Enterprise Generative AI on GCP
In the current epoch of Artificial Intelligence, the industry remains singularly preoccupied with the "Model" — obsessing over the raw parameter sca…
I went looking for my ad profile inside the bid request. It wasn't there.
When an ad is auctioned for you, a bid request goes out that describes you — segments, identity graphs, consent state, location precision. Buyers se…
I found a $45 billion number that vanished from NVIDIA's SEC filings
Liquid syntax error: 'raw' tag was never closed
LAB now ships a free Idea Feed: rule-shaped trading ideas, deliberately untested
A small release, not a launch. The LAB tab on gex.live has a new rightmost rail called IDEA FEED . It is a stream of short, rule-shaped trading idea…
The Lab: a backtester that is allowed to say "no"
gex.live has two halves. The terminal measures where SPX options dealers are positioned, every second, from the tape. The Lab is the half that asks…
What 32 leaked brand names taught me about auditing affiliate templates
I was comparing crypto trading terminals and did what I usually do first: curl the top-ranking page and read the HTML instead of the rendering. The…
Point-in-Time Fundamentals for Numerai Signals: Killing Lookahead in Your Feature Join
If you build features for Numerai Signals from fundamentals, the single most common way to silently overstate your live performance is joining on th…
Checking your A/B test every morning makes it lie to you
You launch the test on Monday. You check the dashboard each morning, because of course you do. On Thursday it reads p = 0.04, you call it, and you s…
Our Documentation Was Lying. The Model Believed It.
There is a variable called slp — sea-level pressure. Our API documentation lists it as available. Our error messages list it among the valid options…
Why Extracting Tables From a PDF Is Harder Than It Looks (and How We Actually Do It)
If you have ever copy-pasted a table out of a PDF, you already know what happens. Rows collapse into one long line of text. Columns interleave. Numb…
24 of the 30 duplicate groups were not duplicates
We grouped an asset database by title and found 30 groups of "duplicates." Content hashing dissolved 24 of them. Five were real. One was unjudgeable…
Getting Started with WEKA: A Beginner’s Guide to Machine Learning Without Code
Getting started with machine learning WEKA for Beginners: A Practical Introduction to Machine Learning Without Code Getting started with machine lea…
Multi-Armed Bandit Testing: How It Works and When to Use
A multi-armed bandit is an algorithm that decides how to split traffic across several variations while a test is still running, shifting more visito…