📚 Data Science Riddle
What metric is commonly used to decide splits in decision trees?
What metric is commonly used to decide splits in decision trees?
Anonymous Quiz
56%
Entropy
18%
Accuracy
6%
Recall
20%
Variance
❤4
📚 Data Science Riddle
Which algorithm is most sensitive to feature scaling?
Which algorithm is most sensitive to feature scaling?
Anonymous Quiz
25%
Decision Tree
24%
Random Forest
36%
KNN
15%
Naive Bayes
📚 Data Science Riddle
Why does bagging reduce variance?
Why does bagging reduce variance?
Anonymous Quiz
13%
Uses deeper trees
50%
Averages multiple models
29%
Penalizes weights
9%
Learns Sequentially
📊 Infographic Elements That Every Data Person Should Master 🚀
After years of working with data, I can tell you one thing:
👉 The
Here’s your quick visual toolkit 👇
🔹 Timelines
* Sequential ⏩ great for processes
* Scaled ⏳ best for real dates/events
🔹 Circular Charts
* Donut 🍩 & Pie 🥧 for proportions
* Radial 🌌 for progress or cycles
* Venn 🎯 when you want to show overlaps
🔹 Creative Comparisons
* Bubble 🫧 & Area 🔵 for impact by size
* Dot Matrix 🔴 for colorful distributions
* Pictogram 👥 when storytelling matters most
🔹 Classic Must-Haves
* Bar 📊 & Histogram 📏 (clear, reliable)
* Line 📈 for trends
* Area 🌊 & Stacked Area for the “big picture”
🔹 Advanced Tricks
* Stacked Bar 🏗 when categories add up
* Span 📐 for ranges
* Arc 🌈 for relationships
💡 Pro tip from experience:
If your audience doesn’t “get it” in 3 seconds, change the chart. The best visualizations
After years of working with data, I can tell you one thing:
👉 The
chart ou choose is as important as the data itself.Here’s your quick visual toolkit 👇
🔹 Timelines
* Sequential ⏩ great for processes
* Scaled ⏳ best for real dates/events
🔹 Circular Charts
* Donut 🍩 & Pie 🥧 for proportions
* Radial 🌌 for progress or cycles
* Venn 🎯 when you want to show overlaps
🔹 Creative Comparisons
* Bubble 🫧 & Area 🔵 for impact by size
* Dot Matrix 🔴 for colorful distributions
* Pictogram 👥 when storytelling matters most
🔹 Classic Must-Haves
* Bar 📊 & Histogram 📏 (clear, reliable)
* Line 📈 for trends
* Area 🌊 & Stacked Area for the “big picture”
🔹 Advanced Tricks
* Stacked Bar 🏗 when categories add up
* Span 📐 for ranges
* Arc 🌈 for relationships
💡 Pro tip from experience:
If your audience doesn’t “get it” in 3 seconds, change the chart. The best visualizations
speak louder than numbers❤8🔥3
📚 Data Science Riddle
Which Metric is best for imbalanced classification?
Which Metric is best for imbalanced classification?
Anonymous Quiz
20%
Accuracy
18%
Precision
18%
Recall
44%
F1-Score
📚 Data Science Riddle
A dataset has 20% missing values in a critical column. What's the most practical choice?
A dataset has 20% missing values in a critical column. What's the most practical choice?
Anonymous Quiz
6%
Drop all rows
49%
Fill with mean/median
41%
Use model-based imputation
5%
Ignore missing data
❤3
ML models don’t all think alike 🤖
❇️ Naive Bayes = probability
❇️ KNN = proximity
❇️ Discriminant Analysis = decision boundaries
Different paths, same goal: accurate classification.
Which one do you reach for first?
❇️ Naive Bayes = probability
❇️ KNN = proximity
❇️ Discriminant Analysis = decision boundaries
Different paths, same goal: accurate classification.
Which one do you reach for first?
❤4
📚 Data Science Riddle
In a medical diagnosis project, what's more important?
In a medical diagnosis project, what's more important?
Anonymous Quiz
34%
High precision
15%
High recall
37%
High accuracy
14%
High F1-score
Important LLM Terms
🔹 Transformer Architecture
🔹 Attention Mechanism
🔹 Pre-training
🔹 Fine-tuning
🔹 Parameters
🔹 Self-Attention
🔹 Embeddings
🔹 Context Window
🔹 Masked Language Modeling (MLM)
🔹 Causal Language Modeling (CLM)
🔹 Multi-Head Attention
🔹 Tokenization
🔹 Zero-Shot Learning
🔹 Few-Shot Learning
🔹 Transfer Learning
🔹 Overfitting
🔹 Inference
🔹 Language Model Decoding
🔹 Hallucination
🔹 Latency
🔹 Transformer Architecture
🔹 Attention Mechanism
🔹 Pre-training
🔹 Fine-tuning
🔹 Parameters
🔹 Self-Attention
🔹 Embeddings
🔹 Context Window
🔹 Masked Language Modeling (MLM)
🔹 Causal Language Modeling (CLM)
🔹 Multi-Head Attention
🔹 Tokenization
🔹 Zero-Shot Learning
🔹 Few-Shot Learning
🔹 Transfer Learning
🔹 Overfitting
🔹 Inference
🔹 Language Model Decoding
🔹 Hallucination
🔹 Latency
❤11