Kelvin Genao Carmona

Freelance Data Scientist & Data Analyst | SQL, Python, Power BI
🚀 Making data speak | 📊 Insights that drive decisions

[Kelvin Genao Carmona @KelvinGenaoC]

I’m a freelance Data Scientist & Analyst sharing:

📌 DS TIPS (tools, code, workflows)
📌 TECH & DATA NEWS
📌 Projects & real insights
If you’re into Python, ML, or data — follow along. Let’s grow together 👇

[Kelvin Genao Carmona @KelvinGenaoC]

Always check for data leakage before training your ML model.
It can inflate your accuracy and make your model useless in the real world.
🔍 Split your data first → then preprocess.
🔥 Never use test data during feature engineering.

#DataScience #MachineLearningTips

[Kelvin Genao Carmona @KelvinGenaoC]

Microsoft strengthens its AI leadership with key partnerships at Build 2025. New models and digital assistants arrive on Azure.
🔗 https://t.co/3Q8UOnIIY9
#AInews #MicrosoftBuild

[Kelvin Genao Carmona @KelvinGenaoC]

Microsoft’s new AI model Aurora is changing the weather forecasting game — predicting cyclone paths better than traditional methods and using less computing power. A big step forward in climate tech. 🌪️🌍
Source: https://t.co/EbHpqcsYWP
#TechNews #AI #DataScience

[Kelvin Genao Carmona @KelvinGenaoC]

But hey, I’m not just here to complain...
Follow me!
In the next post, I’ll show you how to clean data faster—without spending hours searching for errors manually.
And if you’ve ever faced a horrifying Excel file…
Drop a “That’s me!” or share your worst cleaning nightmare 👇

[Kelvin Genao Carmona @KelvinGenaoC]

Why do analysts, data scientists, and anyone who works with data complain so much about data cleaning?
Isn’t it just removing nulls and duplicates? 🤔
🧵Thread:

[Kelvin Genao Carmona @KelvinGenaoC]

So yes, we complain.
Because data cleaning is hard, slow, and critical.
If you do it wrong, your whole analysis falls apart.

[Kelvin Genao Carmona @KelvinGenaoC]

🧮 Why does your model perform worse on new data?
It’s probably overfitting.
The model learned the noise, not the pattern.
Tip:
✅ Use cross-validation
✅ Simplify features
✅ Regularize
A good model predicts, not memorizes.
#MLTips #Overfitting

[Kelvin Genao Carmona @KelvinGenaoC]

🧮 Why does your model perform worse on new data?
It’s probably overfitting.
The model learned the noise, not the pattern.
Tip:
✅ Use cross-validation
✅ Simplify features
✅ Regularize
A good model predicts, not memorizes.
#MLTips #Overfitting

[Kelvin Genao Carmona @KelvinGenaoC]

💡 Learn this and you'll stand out as an analyst:
👉 Ask better questions.
"What happened?" is easy.
"Why did it happen?" and "What will happen?" make you great.
Insights come from questions, not dashboards.
#DataAnalytics #Storytelling

[Kelvin Genao Carmona @KelvinGenaoC]

📈 Correlation ≠ Causation
Just because two variables move together doesn’t mean one causes the other.
Example: Ice cream sales & drowning incidents rise in summer.
Are they related? Yes.
Does one cause the other? No.
Always dig deeper.
#DataScience #Stats

[Kelvin Genao Carmona @KelvinGenaoC]

🔍 Want to improve your data analysis?
Before using fancy models, do this:
✅ Remove duplicates
✅ Handle missing values
✅ Check for outliers
✅ Analyze correlations
Great analysis starts with clean data.
Cleaning beats modeling, every time.
#DataScience #Python

[Kelvin Genao Carmona @KelvinGenaoC]

🔥 Quick SQL trick:
Want to count how many unique customers bought each month?
SELECT MONTH(date), COUNT(DISTINCT customer_id)
FROM sales
GROUP BY MONTH(date);

More useful than 10 Excel functions 😉
#SQL #DataTips

[Kelvin Genao Carmona @KelvinGenaoC]

🔍 Unsupervised Learning isn’t about predicting — it’s about discovering.
Clustering & Dimensionality Reduction help us explore data, find hidden patterns, and simplify complexity.
A great starting point for understanding your data 👇
#MachineLearning #Clustering #Unsupervised