Tag feed

#datascience

171 posts29 followers

Explore Hashnode

Alternatives

Trending tags this week

FMFaith Mwendebeginnerslog.hashnode.dev3d ago · 3 min read

Splitting Fraud Data Without Breaking the Ratio (Day 3)

Day 1 was setup. Day 2 was looking at the data. Day 3 was the first real modeling decision: how do I split this data without breaking it? The problem it solves You can't train a model on all your data

0

FMFaith Mwendebeginnerslog.hashnode.dev5d ago · 3 min read

What I Learned From My First Look at Fraud Data (Day 2)

Day 1 was setup. Day 2 was the first time I actually looked at the data and the numbers immediately broke my assumptions. Here's what the credit card fraud dataset told me. The problem it solves I had

0

MHMichael Hairetisopenred.hashnode.dev6d ago · 5 min read

A Forecasting Model That Correctly Refuses to Forecast

It beat the baseline by 98.7% on one series and lost to repeating the last value on another. That contrast is the most useful single fact about it. Part one covered what TimesFM 3.0 is: a stateless n

1

S

VBVivek Biswaszero2ds.hashnode.devSep 21 · 30 min read

Days 6-14: From Corporate Actions to My First Full Annual Report, the Complete Week 2 Recap

Day 6: Asset Turnover, Cash Cycles, and What Corporate Actions Actually Do to Your Shares Today was a concept-heavy day, four new ideas back to back, and for the first time I actually needed a table t

0

VBVivek Biswaszero2ds.hashnode.devSep 11 · 4 min read

Days 2-5: A Missed Morning, Gradebook Charts, and My First Real Fundamental Analysis Concepts

Catching up on four days at once here, since life did not exactly cooperate with the schedule this week. Day 2: When the plan meets real life Started strong: up at 5:30 AM, a proper workout, and a 4 k

0

MMohitmmonthlydatascience.hashnode.devAug 23 · 7 min read

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, drop one variable from every pair above 0.8. Then we

0

PPetrichormeteosource.hashnode.devAug 18 · 6 min read

Build a wind generation forecast dashboard in Python

A wind farm's output has nothing to do with the clock and everything to do with the weather. Unlike solar, which follows a predictable daily curve, wind generation can spike at 2am and drop to almost

0

BBBahati Brenda Kizitogodfident-data.hashnode.devAug 16 · 6 min read

I Tried Predicting House Prices… Then scikit-learn Gave Me a History Lesson

I just wanted to predict house prices. I opened a notebook, typed the familiar import, and wrote the line I had seen in dozens of tutorials: from sklearn.datasets import load_boston boston = load_bost

0

BBBahati Brenda Kizitogodfident-data.hashnode.devAug 15 · 8 min read

99% Accuracy Is Not Always Good: What Fraud Detection Taught Me About ML

I built a fraud detection model that hit 99% accuracy. I was proud of it for about twelve minutes. Then I looked closer and realized the model was basically doing nothing useful. This is the story of

0

GGTZHostgtzhost.hashnode.devAug 7 · 2 min read

Architectural Breakdown: NVIDIA H100, A100, and L40S for Modern Workloads

Choosing the right Tensor Core GPU for a data center deployment requires understanding the architectural nuances of modern NVIDIA chips. Throwing hardware at a problem without analyzing the specific w

0

#datascience

Search Hashnode

#datascience

Explore Hashnode

Trending tags this week

Splitting Fraud Data Without Breaking the Ratio (Day 3)

What I Learned From My First Look at Fraud Data (Day 2)

A Forecasting Model That Correctly Refuses to Forecast

Days 6-14: From Corporate Actions to My First Full Annual Report, the Complete Week 2 Recap

Days 2-5: A Missed Morning, Gradebook Charts, and My First Real Fundamental Analysis Concepts

Correlation Is Pairwise. Multicollinearity Isn't

Build a wind generation forecast dashboard in Python

I Tried Predicting House Prices… Then scikit-learn Gave Me a History Lesson

99% Accuracy Is Not Always Good: What Fraud Detection Taught Me About ML

Architectural Breakdown: NVIDIA H100, A100, and L40S for Modern Workloads