Homework Objective
The original pipelines processed NYC Taxi data for **2019 and 2020**.
The task was to **extend the existing flows to include data for 2021**, specifically:
- **January 2021 – July 2021**
- **Both Yellow and Green Taxi datasets*
I am pursuing Data- Engineering Zoomcamp from Datat talk club : what I learned in the first week
The homework covers:
\- Running Docker containers
\- Setting up PostgreSQL and pgAdmin with Docker Compose
\- Loading NYC Green Taxi data into Postgres
\- Writing SQL queries to answer analytical questions
\- Basic Terraform workflow concepts
0 likes, 0 comments - riniplantmom on November 1, 2025: "In this chapter of the ML Zoomcamp by DataTalks.Club (led by Alexey Grigorev), we dived into Decision Trees and Ensemble Learning—two core components in supervised machine learning that offer high interpretability and flexibility. This chapter addresses decision trees, their structure, splitting methods, as well as ensemble techniques like bagging, boosting, and stacking to improve model performance. Notable briefings on the same are as follows: Decision Trees: Core Concepts and Learning In this section, the course covers decision trees as intuitive, rule-based algorithms that are effective yet prone to overfitting on complex datasets. Key topics include: Splitting Criteria: Decision trees divide data by optimizing splits to minimize classification error. Concepts like "impurity" are introduced, helping learners understand how criteria such as Gini impurity and entropy guide the algorithm in choosing splits that reduce classification mistakes. Overfitting risks are discussed, particularly with deep trees that may learn too much noise from training data. Hyperparameters Tuning: Overfitting risks are addressed through hyperparameters like max_depth and min_samples_split, which limit the tree’s depth or require a minimum number of data points to create a split. This control helps maintain model generalizability. Random Forests: Reducing Variance with Bagging Reduce Variance: By training multiple trees on bootstrapped samples and averaging their predictions, Random Forests minimize the variance seen in individual decision trees. Each tree votes, and the most common prediction is taken as the final output. Feature Randomization: Not only are data samples randomized, but each split only considers a random subset of features, reducing correlation among trees and further lowering overfitting risks. Hyperparameters Tuning: Important parameters include n_estimators (number of trees) and max_features (maximum features per split). Tuning these parameters helps balance model performance and computational cost, which is demonstrated through hands-on coding examples in Python. Boosting: Correcting Weak Learners Boosting techniques improve model accuracy by correcting".
