Ayush @TensorThrottleX

🤖 AI/ML enthusiast | 📊 Data whisperer in training | 📚 DSA wrangler | 🧠 Teaching machines to think (kinda) | ⚙️ Compiler’s fault. Always. #100daysofcode nowhere Joined April 2025

Tweets

715
Followers

305
Following

334
Likes

3K

Ayush @TensorThrottleX

3 hours ago

Day 111 – Data Science Journey -> WCSS: cluster compact, sum of sq.dist. -> Elbow Method: WCSS vs k, optimal clusters -> EM Algo: soft clust, prob approach -> GMM: Gaussian Mixture Models with EM optimization #100DaysOfCode #DataScience #MachineLearning #Clustering

0 0 8 77 0

Download Image

Ayush @TensorThrottleX

a day ago

Day 110 – Data Science Journey -> K-Means Clustering :variable dataset -> Raw data points with scatter plots -> KMeans (n_clusters=3) and assigned cluster labels -> Plotted clusters with centroids for clear separation #100DaysOfCode #DataScience #MachineLearning #Clustering

4 0 16 2K 2

Download Image

Ayush @TensorThrottleX

2 days ago

Day 109 – Data Science Journey -> K-Means Clustering: Partitioning n points into k clusters -> Objective: Minimize intra-cluster variance -> Complexity: O(nkdt) -> Applications: Segmentation, Customer Analytics #100DaysOfCode #DataScience #MachineLearning

1 0 22 858 1

Download Image

Ayush @TensorThrottleX

3 days ago

Day 108 – Data Science Journey -> End-to-End ML Pipeline Creation -> Numeric: Imputer + StandardScaler -> Categorical: Imputer + OneHotEncoder -> Logistic Regression & Random Forest Test -> Saved + Reloaded Models :Joblib for Reuse #100DaysOfCode #DataScience #MachineLearning #AI

5 0 21 853 3

Download Image

Ayush @TensorThrottleX

4 days ago

Day 107 – Data Science Journey -> Built a basic ML Pipeline -> One-Hot Encoding + MinMax Scaling -> RFECV (RandomForest) - Optimal Features -> Model Eval & Pipeline -> Precision–Recall Threshold Tuning #100DaysOfCode #DataScience #MachineLearning

8 1 19 645 1

Download Image

Ayush @TensorThrottleX

5 days ago

Day 106 – Data Science Journey -> Grid Search – Hyperparameter Tuning -> Parameter combos (exhaustive search) -> Eg: n_estimators, max_depth in RF -> Finds best model settings automatically -> Costly for big grids, gives optimal result #100DaysOfCode #DataScience #MachineLearning

0 1 16 750 4

Download Image

Ayush @TensorThrottleX

6 days ago

Day 105 – Data Science Journey -> KNN Pipeline Implementation -> One-Hot Encoding + MinMax Scaling -> RFECV (RandomForest): Optimal Features -> Trained & Evaluated with Confusion Matrix + Metrics -> Tuned thresholds: Precision–Recall #100DaysOfCode #DataScience #MachineLearning

1 2 13 2K 2

Download Image

~csy😎 @the__csy20

7 days ago

My bro Om just dropped his new track Kaise Rahun 🎶 Been vibin, this one really hits different. Check it out 👉 youtu.be/8ND42yqU85Q?si…

5 11 16 4K 1

Ayush @TensorThrottleX

a week ago

Day 104 – Data Science Journey ->KNN – Choosing the Right k -> Small k (1–2): noisy, unstable -> Large k: smooth, biased ->Odd k: avoids ties (binary case) -> Plot k vs accuracy = find the sweet spot #100DaysOfCode #DataScience #MachineLearning

1 0 18 2K 0

Download Image

Ayush @TensorThrottleX

a week ago

Day 103 – Data Science Journey -> KNN Classification with Euclidean distance -> Feature Scaling: Standardization & Normalization -> Derived scaling equations -> Compared performance on unscaled vs standardized vs normalized data #100DaysOfCode #DataScience #MachineLearning

1 0 14 2K 2

Download Image

Ayush @TensorThrottleX

a week ago

Day 102 – Data Science Journey ->Learned KNN Classifier on balanced dataset ->Performed train-test split, model fitting & predictions ->Evaluated model using accuracy score ->Analyzed impact of feature selection on performance #100DaysOfCode #DataScience #MachineLearning

5 1 17 575 1

Download Image

Ayush @TensorThrottleX

a week ago

Day 101 – Data Science Journey ->Random Forest Classifier on synthetic dataset ->Implemented fitting, predictions & permutation importance ->Evaluated with confusion matrix & accuracy score ->Feature reduction & contribution- shuffling #100DaysOfCode #DataScience #MachineLearning

7 1 19 731 1

Download Image

Ayush @TensorThrottleX

2 weeks ago

Day 100 – Data Science Journey -> Random Forest Classifier on synthetic “Profit” dataset -> Explored train-test split, model fitting, predictions -> Evaluated with confusion matrix & accuracy score -> generated synthetic datasets #100DaysOfCode #DataScience #MachineLearning

6 0 13 590 0

Download Image

Ayush @TensorThrottleX

2 weeks ago

Day 99 – Data Science Journey -> Data cleaning (dropna, IQR outliers) -> RFECV for feature selection -> Decision Tree with max_depth tuning -> Metrics: Accuracy, Precision, Recall, F1 -> Visualized tree & threshold trade-offs #100DaysOfCode #DataScience #MachineLearning

2 0 15 480 1

Download Image

Ayush @TensorThrottleX

2 weeks ago

Day 98 – Data Science Journey -> Formula: Gini=1−∑(pi)^2 where pi= class probability -> Value range: 0 (pure node) → max (high impurity) -> Used in Decision Trees (CART) to find the best split -> Goal: minimize Gini after splitting #100DaysOfCode #DataScience #MachineLearning

2 0 21 550 2

Download Image

Ayush @TensorThrottleX

2 weeks ago

Day 97 – Data Science Journey ->DecisionTreeClassifier: classification tasks ->Train-test split, model training, acc_evaluation -> Overfitting : detected by comparing train vs test accuracy -> Visualizing decision trees with plot_tree #100DaysOfCode #DataScience #MachineLearning

3 0 19 573 5

Download Image

Ayush @TensorThrottleX

2 weeks ago

Day 96 – Data Science Journey -> Preprocessing: missing vals, outliers -> FeatureSel: RFECV + LogReg -> Modeling: Logistic Regression -> Eval: ConfMatrix, Acc, Prec, Rec, F1 -> Threshold: Prec–Rec–F1 tradeoff #100DaysOfCode #DataScience #MachineLearning

4 0 18 800 1

Download Image

Ayush @TensorThrottleX

2 weeks ago

Day 95 – Data Science Journey -> FPR: negatives wrongly predicted as positives -> TPR (Recall/Sensitivity): measures positives captured -> ROC Curve: TPR vs FPR across thresholds -> Logist. Reg vs RandomForest: compared ROC performance #100DaysOfCode #DataScience #MachineLearning

0 0 12 266 1

Download Image

Ayush @TensorThrottleX

2 weeks ago

Day 94 – Data Science Journey 🔹 Classification Accuracy : a quick measure of performance often misleading 🔹 Confusion Matrix : way to visualize true +ves, TN, FN, FP 🔹 Advanced evaluation metrics : Precision, Recall, Sensitivity #100DaysOfCode #DataScience #MachineLearning

7 0 21 835 4

Download Image

Ayush @TensorThrottleX

3 weeks ago

🚀 Day 93 – Data Science Journey 🔹 Logistic Regression: Visualizing decision boundaries, prob curves 🔹 Logistic Regression eq & decision boundary 🔹Confusion matrix :classification tasks 🔹 Sigmoid Curve:probabilities between 0 and 1 #100DaysOfCode #DataScience #MachineLearning