subair9
subair9
Untitled
24 posts
Don't wanna be here? Send us removal request.
subair9 · 4 months ago
Text
Article
Just published an article on Using Bagging And XGBoost To Train Large Datasets. The link to the article is shown below.
0 notes
subair9 · 5 months ago
Text
Tumblr media
Just completed the ML Zoomcamp 2024 Competition on kaggle which is a Retail Demand Forecasting Challenge! The competition challenges participants to forecast customer demand based on historical sales and related data from a retailer.
The dataset consists of 10 files in csv format with a total size of 872.28 MB and the files had to be aggregated to form a single dataset of 7,432,685 rows, and 16 columns.
Creating a model for this dataset was a challenge because the size was to large for all the platforms I used so, I have to apply an ensemble method where a fraction of the date was modelled in series and later combined at the end.
This was a wonderful experience that was much more different from the toy dataset that we are used to.
The link to the competition is below: https://www.kaggle.com/competitions/ml-zoomcamp-2024-competition/overview
0 notes
subair9 · 5 months ago
Text
Tumblr media
ML Zoomcamp Project Capstone 3 – 3: First trainings
After conducting the exploratory data analysis, 3 pre-trained deep convolutional neural network models were used to train the data for 10 epochs each with the aim of finding the one that will return the best test accuracy which will then be used to create the final model. The results obtained with BATCH_SIZE = 32, LEARNING_RATE = 0.0001 are shown below:
Xception - 98.93%
InceptionV3 - 98.32%
ResNet101V2 - 96.80%
Xception was selected because of its higher test accuracy. Attached is the full output of the trainings.
0 notes
subair9 · 5 months ago
Text
Tumblr media
ML Zoomcamp Project Capstone 2 – 2: Exploratory Data Analysis
The dataset used in this project is the Brain Tumor MRI Dataset https://www.kaggle.com/datasets/masoudnickparvar/brain-tumor-mri-dataset from kaggle.
It consists of 7,023 magnetic resonance imaging scans, annotated in a folder structure of 5,712 test and 1,311 train. The dataset consists of images of no tumour and brain tumor types: pituitary, meningioma, and glioma, as shown in the attached diagram.
0 notes
subair9 · 5 months ago
Text
ML Zoomcamp Project Capstone 2 – 1: Problem Statement
A brain tumor is a mass of abnormal cells that grows in the brain, it can be benign (non-cancerous) or malignant (cancerous). Its symptoms are:
Headaches, which can be severe, persistent, or come and go
Seizures, which can be mild or severe
Weakness or paralysis in part of the body
Loss of balance
Changes in mood or personality
Changes in vision, hearing, smell, or taste
Nausea and vomiting
Difficulty speaking
Difficulty swallowing
The growth of brain tumors can cause the pressure inside the skull to increase leading to brain damage, and loss of life if not discovered early and properly treated.
This project is aimed at developing a robust brain tumor detection model using Convolutional Neural Networks (CNNs) to automate the analysis of magnetic resonance imaging (MRI) scan by accurately identifying and classifying the tumors at an early stage which can reduce the load on doctors, help in selecting the most convenient treatment method and hence increase the rate of survival.
0 notes
subair9 · 6 months ago
Text
ML Zoomcamp
Just completed the tenth week of Machine Learning Zoomcamp.
The lessons covered include:
TensorFlow Serving
Creating a pre-processing service
Running everything locally with Docker-compose
Introduction to Kubernetes
Deploying a simple service to Kubernetes
Deploying TensorFlow models to Kubernetes
Deploying to EKS
The link to the course is below: https://github.com/DataTalksClub/machine-learning-zoomcamp
0 notes
subair9 · 6 months ago
Text
Machine Learning Zoomcamp
Just completed the ninth week of Machine Learning Zoomcamp.
The lessons covered include:
Introduction to Serverless
AWS Lambda
TensorFlow Lite
Preparing the code for Lambda
Preparing a Docker image
Creating the lambda function
API Gateway: exposing the lambda function
The link to the course is below: https://github.com/DataTalksClub/machine-learning-zoomcamp
1 note · View note
subair9 · 7 months ago
Text
Machine Learning Zoomcamp
Just completed the eighth week of Machine Learning Zoomcamp.
The lessons covered include:
Fashion classification
Setting up the Environment on Saturn Cloud
TensorFlow and Keras
Pre-trained convolutional neural networks
Convolutional neural networks
Transfer learning
Adjusting the learning rate
Checkpointing
Adding more layers
Regularization and dropout
Data augmentation
Training a larger model
Using the model
The link to the course is below: https://github.com/DataTalksClub/machine-learning-zoomcamp
1 note · View note
subair9 · 8 months ago
Text
ML Zoomcamp
Just completed the sixth week of Machine Learning Zoomcamp.
The lessons covered include:
Credit risk scoring project
Data cleaning and preparation
Decision trees
Decision tree learning algorithm
Decision trees parameter tuning
Ensemble learning and random forest
Gradient boosting and XGBoost
XGBoost parameter tuning
Selecting the best model
The link to the course is below: https://github.com/DataTalksClub/machine-learning-zoomcamp
1 note · View note
subair9 · 8 months ago
Text
ML Zoomcamp
Just completed the fifth week of Machine Learning Zoomcamp.
The lessons covered include:
Saving and loading the model
Web services: introduction to Flask
Serving the churn model with Flask
Python virtual environment: Pipenv
Environment management: Docker
Deployment to the cloud: AWS Elastic Beanstalk
The link to the course is below: https://github.com/DataTalksClub/machine-learning-zoomcamp
0 notes
subair9 · 8 months ago
Text
LLM Zoomcamp
Just submitted my final project for LLM Zoomcamp 2024. It has been a wonderful experience because the course gave me the opportunity to work with both closed and open source LLMs. The knowledge attained include vector-search, monitoring, orchestration and best practices. My sincere gratitude goes to Alexey Grigorev for making all these possible.
The link to the project is below:
0 notes
subair9 · 8 months ago
Text
ML Zoomcamp
Just completed the forth week of Machine Learning Zoomcamp.
The lessons covered include: 4.1 Evaluation metrics: session overview 4.2 Accuracy and dummy model 4.3 Confusion table 4.4 Precision and Recall 4.5 ROC Curves 4.6 ROC AUC 4.7 Cross-Validation
The link to the course is below: https://github.com/DataTalksClub/machine-learning-zoomcamp
2 notes · View notes
subair9 · 8 months ago
Text
ML Zoomcamp
Just completed the third week of Machine Learning Zoomcamp.
The lessons covered include: 3.1 Churn prediction project 3.2 Data preparation 3.3 Setting up the validation framework 3.4 EDA 3.5 Feature importance: Churn rate and risk ratio 3.6 Feature importance: Mutual information 3.7 Feature importance: Correlation 3.8 One-hot encoding 3.9 Logistic regression 3.10 Training logistic regression with Scikit-Learn 3.11 Model interpretation 3.12 Using the model 3.13 Summary
The link to the course is below: https://github.com/DataTalksClub/machine-learning-zoomcamp
1 note · View note
subair9 · 9 months ago
Text
ML Zoomcamp
Just completed the second week of Machine Learning Zoomcamp.
The lessons covered include: 1 Car price prediction project 2 Data preparation 3 Exploratory data analysis 4 Setting up the validation framework 5 Linear regression 6 Linear regression: vector form 7 Training linear regression: Normal equation 8 Baseline model for car price prediction project 9 Root mean squared error 10 Using RMSE on validation data 11 Feature engineering 12 Categorical variables 13 Regularization 14 Tuning the model 15 Using the model 16 Car price prediction project summary 17 Explore more
The link to the course is below: https://github.com/DataTalksClub/machine-learning-zoomcamp
2 notes · View notes
subair9 · 9 months ago
Text
ML Zoomcamp
Just completed the first week of Machine Learning Zoomcamp.
The lessons covered include:
What is Machine Learning
Machine Learning vs Rules Based Systems
Supervised Machine Learning
CRISP-DM
Model Selection Process
Setting up the Environment
Introduction to NumPy
Linear Algebra Refresher
Introduction to Pandas
Summary
The link to the course is below: https://github.com/DataTalksClub/machine-learning-zoomcamp
1 note · View note
subair9 · 10 months ago
Text
LLM Zoomcamp
Just completed the fifth week of LLM Zoomcamp.
The lessons covered include:
Review
Ingest
Chunk
Tokenization
Embed
Export
Retrieval
Trigger Daily Runs
The link to the course is below: https://github.com/DataTalksClub/llm-zoomcamp
1 note · View note
subair9 · 11 months ago
Text
LLM Zoomcamp 2004
Just completed the forth week of LLM Zoomcamp.
The lessons covered include:
Introduction to monitoring answer quality
Evaluation and Monitoring in LLMs
Offline RAG Evaluation
Offline RAG Evaluation: Cosine Similarity
Offline RAG Evaluation: LLM as a Judge
Capturing User Feedback
Monitoring the System
The link to the course is below: https://github.com/DataTalksClub/llm-zoomcamp
0 notes