Pinned Tweet
This November, let's focus on "N":
> No excuses
> New habits
> No days wasted
> Nutrition
> No distractions
> Never settling
NOBODY IS COMING TO SAVE YOU 🫵
Saumya retweeted
bhagwan kare ki sabka accha placement ho jaaye yaar!
deserving logo ka nhi hota toh bahut bura feel hota hai kasam se
Glad that when I got the offer nobody called it diversity hiring or luck based 🥰 but instead they said you deserve better😭
This quoted post is unavailable.
7
16
691
The only article you need to get started with System Design.
Thank me later
medium.com/@shivambhadani_/s…
8
258
Saumya retweeted
Got featured with the @vg19988 🐐
Sound quality was very bad ik 😔
SUPERVISED LEARNING ROADMAP
│
├── 1. Fundamentals
│ ├── Introduction to Supervised Learning
│ ├── Regression vs Classification
│ ├── Problem Formulation: Features (X) and Labels (Y)
│ ├── Model Fitting
│ │ ├── Underfitting (High Bias)
│ │ ├── Overfitting (High Variance)
│ │ └── Bias Variance Tradeoff
│ ├── Generalization
│ ├── Training, Validation & Test Sets
│ └── Data Leakage
│
├── 2. Regression Algorithms
│ ├── Linear Regression
│ │ ├── Ordinary Least Squares (OLS)
│ │ ├── Model Assumptions
│ │ ├── Residual Analysis
│ │ └── Regularization
│ │ ├── L1 (Lasso)
│ │ ├── L2 (Ridge)
│ │ └── Elastic Net
│ └── Polynomial Regression [Problem Dependent]
│
├── 3. Classification Algorithms
│ ├── Logistic Regression
│ │ ├── Sigmoid Function
│ │ ├── Decision Boundary
│ │ ├── Classification Threshold
│ │ └── Regularization
│ ├── Naive Bayes
│ ├── K Nearest Neighbors (KNN)
│ └── Support Vector Machines (SVM)
│ └── Kernel Trick
│
├── 4. Loss Functions
│ ├── Regression Loss
│ │ ├── Mean Squared Error (MSE)
│ │ ├── Mean Absolute Error (MAE)
│ │ └── Huber Loss
│ └── Classification Loss
│ ├── Binary Cross Entropy
│ ├── Categorical Cross Entropy
│ └── Hinge Loss
│
├── 5. Optimization Algorithms [DL]
│ ├── Closed Form Optimization
│ │ └── Normal Equation
│ └── Gradient Descent
│ ├── Batch Gradient Descent
│ ├── Stochastic Gradient Descent
│ ├── Mini Batch Gradient Descent
│ └── Advanced Optimizers
│ ├── Adagrad
│ ├── RMSProp
│ ├── Adam
│ ├── AdamW
│ ├── Muon [Advanced]
│ └── SOAP [Advanced]
│
├── 6. Decision Trees & Ensemble Learning
│ ├── Decision Trees
│ │ ├── Recursive Feature Partitioning
│ │ ├── Gini Impurity
│ │ ├── Entropy
│ │ └── Information Gain
│ └── Ensemble Methods
│ ├── Bagging
│ │ └── Random Forest
│ ├── Boosting
│ │ ├── AdaBoost
│ │ ├── Gradient Boosting
│ │ ├── XGBoost
│ │ ├── LightGBM
│ │ └── CatBoost [Advanced]
│ └── Bagging vs Boosting
│
├── 7. Model Evaluation
│ ├── Classification Metrics
│ │ ├── Confusion Matrix (TP, TN, FP, FN)
│ │ ├── Accuracy
│ │ ├── Precision
│ │ ├── Recall
│ │ ├── Specificity
│ │ ├── F1 Score
│ │ ├── Balanced Accuracy
│ │ ├── ROC AUC
│ │ ├── PR AUC
│ │ └── Log Loss
│ └── Regression Metrics
│ ├── MAE
│ ├── MSE
│ ├── RMSE
│ └── R² Score
│
├── 8. Advanced Classification Concepts
│ ├── Multi Class Classification
│ │ ├── One vs Rest (OvR)
│ │ ├── One vs One (OvO)
│ │ └── Softmax Regression
│ ├── Class Imbalance [Problem Dependent]
│ │ ├── SMOTE
│ │ ├── Random Undersampling
│ │ └── Class Weighted Learning
│ └── Probability Calibration [Advanced]
│ ├── Calibration Methods
│ │ ├── Platt Scaling
│ │ └── Isotonic Regression
│ └── Calibration Evaluation Metrics
│ ├── Brier Score
│ └── Expected Calibration Error (ECE)
│
├── 9. Model Selection & Validation
│ ├── Cross Validation
│ │ ├── K Fold Cross Validation
│ │ ├── Stratified K Fold
│ │ ├── Time Series Split [Problem Dependent]
│ ├── Hyperparameter Tuning
│ │ ├── Grid Search
│ │ ├── Random Search
│ │ └── Bayesian Optimization [Advanced]
│ ├── Model Selection
│ └── Learning & Validation Curves
│
└── 10. ML Pipelines
├── Data Preprocessing
│ ├── Missing Value Imputation
│ ├── Numerical Data
│ │ ├── Standardization
│ │ └── Normalization
│ ├── Categorical Data
│ │ ├── One Hot Encoding
│ │ └── Label Encoding
│ └── Outlier Handling
├── Feature Engineering
│ ├── Feature Selection
│ ├── Polynomial Features
│ └── Interaction Features
├── ColumnTransformer
│ └── Handling Mixed Type Data
├── Preprocessing + Model Integration
└── Scikit Learn Pipelines
Remarks:
[Advanced] = can skip or learn after understanding the fundamentals.
[DL] = Can understand in Deep Learning / Neural Networks
[Problem Dependent] = Learn when required by the use case
9
398
Ye naukri lagte hi shaadi ki baat kahase agyi?😭
Isse acha toh berojgaari ke taane sun leti😭
13
42
1,352
Want to understand LLMs from scratch?
Thank me later
medium.com/@saumyayadav213/l…

















