Data Science


- Recent Projects -



Forecasting Stock Price using Prophet

Forecasting Stock Price (time serties data) using Prophet. Workflow forecasting from time series data.

Tools used :
Python, Prophet, yfinance.

Category :
Machine Learning, Time Series, Forecasting



Streaming Tweets with Python and SQLite

Collect a data of streaming tweets in real time. The collected data will be stored in real time also in an SQLite3 database. The code will run on Compute Engine by GCP non stop. This project can provide data scientist to get data realtime. If we can analyse streaming tweets we can understand the feelings of the people about events happening in the present. Business can leverage this information about how the public is reacting to their product. Investors can determine how a stock is going to perform and so much more.

Tools used :
Python, Tweepy, SQLite3, Google Cloud Platform (GCP)

Category :
Data Collecting, Data Engineering, Data Science Stack, GCP.



Supercharge Web Scraping using Asyncio and Python

Scraping Tripadvisor review using asyncronous programming in Python, which is can shorten the execution time. In this project, I compared the synchronous (selenium) and asynchronous (asyncio) scraping methods.

Tools used :
Python, Asyncio, Selenium, Arsenic, Chromedriver, Spacy

Category :
Data Collecting, Web Scraping, Data Science Stack



Article Spinner

For blogger, SEO, digital marketer, website operator, or a writer that always lookout for unique content to post on the website. Article spinning involves creating new and unique articles from an original one.

Tools used :
Python, NLTK

Category :
Unsupervised, NLP.



Data Science for Cybersecurity - Password Strength Meter

A machine learning classification algorithm decide whether our passwords are weak, normal or strong.

Tools used :
Python, pandas, numpy, Scikit-Learn

Category :
Classification, NLP, TF-IDF.



Image Colorization Using Convolutional Neural Network (CNN)

Using deep learning techniques for colorizing grayscale images. By utilizing a pre-train image from Kaggle, we are able to change into CIELAB color space. Then we propose a method to add color into grayscale image using model that we build using Tensorflow. As an application, we colorize the portrait of historical image and obtain interesting result, showing the potential of this method in the growing field of computer-assisted art.

Tools used :
Python, TensorFlow 2.0, Keras.

Category :
Deep Learning, CNN.



Cat and Dog Image Classifier (CNN)

Create Convolutional Neural Network model using TensorFlow and Keras that correctly classifies images of cats and dogs. The project I do are download and explore the dataset, preparing the data, training model using metrics and objective, evaluating and validating the model, and deploying the model. The model is deployed using Flask on Google Cloud Platform (GCP) in the form of a rest API.

Tools used :
Python, TensorFlow 2.0, Keras, Scipy, Pickles, Flask, Postman.

Category :
Classification, Deep Learning, Deploying, GCP.



Simple NLP Project: Hotel Review Sentiment Analysis

Create a Natural Language Processing (NLP) that correctly classify hotel review as either positive or negative statement. The projects I do are download and explore the dataset, prepare the dataset for training, create a model using Scikit-Learn, train the model to fit the data, make predictions using the trained model. The accuracy of the model is 90.9%.

Tools used :
Python, pandas, numpy, matplotlib, Scikit-Learn

Category :
EDA, Classification, NLP.



Book Recommendation Engine using K-Nearest Neighbors

Uses TensorFlow 2.0 and Keras to create a book recommendation algorithm using K-Nearest Neighbors. This project are download and explore the dataset, prepare the dataset for training, create a KNN model using TensorFlow, train the model to fit the data, make predictions using the trained model.

Tools used :
Python, TensorFlow 2.0, Keras.

Category :
Classification, K-Nearest Neighbors.



Linear Regression Health Costs Calculator

Used TensorFlow 2.0 and Keras to predict healthcare costs using a regression algorithm. This project are download and explore the dataset, prepare the dataset for training, create a model using TensorFlow, train the model to fit the data, make predictions using the trained model. The loss of the model is 2787.19.

Tools used :
Python, TensorFlow 2.0, Keras.

Category :
Regression, Machine Learning.



Neural Network SMS Text Classifier (NLP)

Used TensorFlow 2.0 and Keras to create a Natural Language Processing (NLP) that correctly classify SMS messages as either "ham" or "spam". A "ham" message is a normal message sent by a friend. A "spam" message is an advertisement or a message sent by a company. This project are download and explore the dataset, prepare the dataset for training, create an NLP model using TensorFlow, train the model to fit the data, make predictions using the trained model. The accuracy of the model is 98.6%

Tools used :
Python, TensorFlow 2.0, Keras.

Category :
NLP, Classification, Deep Learning.



Insurance Cost Prediction Using Linear Regression

An insurance cost prediction using machine learning with the PyTorch linear regression method. This project are download and explore the dataset, prepare the dataset for training, create a linear regression model, train the model to fit the data, make predictions using the trained model. From the model that I made, the loss obtained was 6353.86.

Tools used :
Python, Pandas, PyTorch

Category :
Regression, Machine Learning.



Classifying Images of Everyday Objects Using a Neural Network

An image classification using the neural network from PyTorch. This project scraping and explore the CIFAR10 dataset, Set up a training pipeline to train a neural network on a GPU, prepare the dataset for training, create a linear regression model, train the model to fit the data, make predictions using the trained model. From the model that I made, the loss obtained was 1.41.

Tools used :
Python, Pandas, PyTorch.

Category :
Classification, Deep Learning.



Weather in Szeged 2006-2016 Prediction Using Linear Regression

A weather in Szeged 2006-2016 prediction using machine learning with the PyTorch linear regression method. The data used is insurance data obtained from Kaggle. The project are download and explore the dataset, prepare the dataset for training, create a linear regression model, train the model to fit the data, make predictions using the trained model. From the model that I made, the loss obtained was 5.70.

Tools used :
Python, Pandas, PyTorch.

Category :
Regression, Machine Learning.

About Me




Data science enthusiast with a strong physics background and experience using Python and R to build statistical modeling, machine learning, data mining, unstructured data analytics to solve challenging business, marketing, and medical problems. Previously I went to Universitas Gadjah Mada. Experienced Contributing Writer about data science and data analytics to Medium Publication. I am a fast learner who likes to challenge myself by participating in data science competitions and active as a competition contributor in Kaggle.