-machine-learning-on-streaming-data-using-kafka-and-docker

I have completed my first project that machine learning on streaming data using Kafka and Docker. You can check-up my GitHub repository for codes.

-voting-bagging-random-forest

Used Python’s Voting and Bagging classifiers along with KNN, logistic regression, decision tree & random forest to study their accuracy. Voting gave the best result.

10.landing-page-fashion

Landing page design using HTML CSS & JS

2019-ncov

Efforts towards proposing a potentially highly active molecule against a target protein of the 2019 Novel Coronavirus

2020_group07_shinyapp

Shiny: Heatmap of proteomics data

3dpointcloudclassification

Challenge to classify 3D point clouds of cities into Ground - Building - Poles - Pedestrians - Cars - Vegetation

ab5_consulting_agritech

Studies on modeling and optimizing the use of sensors, to achieve the best results for farmers. the objective here is to find the best sensor at low cost for smallholders but depending on the type of soil. We use raspberry pi and the 4 channel 16 bit ADC microship for calibration and instrumentation.

ab_testing_final_project_udacity

This is a A/B test project from Udacity.

agriculture-de-decision

airbender

Machine Learning Orchestration with Airflow

airflow_kafka_cassandra_mongodb

Produce Kafka messages, consume them and upload into Cassandra, MongoDB.

algorithmictrading-machinelearning

Deep Learning - Neural network (RNN, LSTM & GRU)

alphaopt

Bayesian Optimisation of Machine Learning Algorithms

anavi-examples

Examples for getting started and testing ANAVI Internet of Things and add-on boards (HAT & pHAT) for Raspberry Pi

andrew-ng-notes

This is Andrew NG Coursera Handwritten Notes.

apache_airflow_pipeline_postgres_to_s3

aquacyder-neeri

The availability of open Earth observation (EO) data through the Copernicus and Landsat programs, as well as plethora of commercially available satellite imagery, represents an unprecedented resource for many EO applications, ranging from ocean and land use/land cover monitoring to disaster control, emergency services and humanitarian relief. Large amounts of such spatiotemporal data call for tools that are able to automatically extract complex patterns embedded inside. eo-learn is a collection of open source Python packages that have been developed to seamlessly access and process spatio-temporal satellite imagery in a timely and automatic manner. It makes the extraction of valuable information from satellite imagery as easy as defining a sequence of operations to be performed on satellite imagery. It also encourages collaboration --- the tasks and workflows can be shared, thus allowing for community-driven ways to exploit EO data. The eo-learn library acts as a bridge between the Earth Observation (EO)/Remote Sensing (RS) field and the Python ecosystem for data science and machine learning. It lowers the entry barrier to the field of RS for non-experts and simultaneously brings the state-of-the-art tools for computer vision, machine learning, and deep learning existing in Python ecosystem to remote sensing experts. AquaCyder aims on tasks like dealing with retrieving the EO data (e.g. Sentinel-2), processing it, adding non-EO data (e.g. labels) to the dataset etc. and finally build the whole pipeline to run such workflow thus preparing the data for ML algorithms for all the water bodies in INDIA, using eo-learn framework

are_you_still_using_elbow_method

argo-iris-pipeline

simple cloud native machine learning pipeline example built with argo workflow and iris dataset

arima

Simple python example on how to use ARIMA models to analyze and predict time series.

atsaf

Applied Time Series Analysis and Forecasting

automobile_data-analysis

Using Python for exploratory data analysis

bank_churn_classification_tensorflow_serving

basic-machine-learning

All the basic ml things

bayesian-regression-and-bitcoin

# Bayesian-Regression-to-Predict-Bitcoin-Price-Variations Predicting the price variations of bitcoin, a virtual cryptographic currency. These predictions could be used as the foundation of a bitcoin trading strategy. To make these predictions, we will have to familiarize ourself with a machine learning technique, Bayesian Regression, and implement this technique in Python. # Datasets We have the datasets in the data folder. The original raw data can be found here: http://api.bitcoincharts.com/v1/csv/. The datasets from this site have three attributes: (1) time in epoch, (2) price in USD per bitcoin, and (3) bitcoin amount in a transaction (buy/sell). However, only the first two attributes are relevant to this project. To make the data to have evenly space records, we took all the records within a 20 second window and replaced it by a single record as the average of all the transaction prices in that window. Not every 20 second window had a record; therefore those missing entries were filled using the prices of the previous 20 observations and assuming a Gaussian distribution. The raw data that has been cleaned is given in the file dataset.csv Finally, as discussed in the paper, the data was divided into a total of 9 different datasets. The whole dataset is partitioned into three equally sized (50 price variations in each) subsets: train1, train2, and test. The train sets are used for training a linear model, while the test set is for evaluation of the model. There are three csv files associated with each subset of data: *_90.csv, *_180.csv, and *_360.csv. In _90.csv, for example, each line represents a vector of length 90 where the elements are 30 minute worth of bitcoin price variations (since we have 20 second intervals) and a price variation in the 91st column. Similarly, the *_180.csv represents 60 minutes of prices and *_360.csv represents 120 minutes of prices. # Project Requirements We are expected to implement the Bayesian Regression model to predict the future price variation of bitcoin as described in the reference paper. The main parts to focus on are Equation 6 and the Predicting Price Change section. # Logic in bitcoin.py 1. Compute the price variations (Δp1, Δp2, and Δp3) for train2 using train1 as input to the Bayesian Regression equation (Equations 6). Make sure to use the similarity metric (Equation 9) in place of the Euclidean distance in Bayesian Regression (Equation 6). 2. Compute the linear regression parameters (w0, w1, w2, w3) by finding the best linear fit (Equation 8). Here you will need to use the ols function of statsmodels.formula.api. Your model should be fit using Δp1, Δp2, and Δp3 as the covariates. Note: the bitcoin order book data was not available, so you do not have to worry about the rw4 term. 3. Use the linear regression model computed in Step 2 and Bayesian Regression estimates, to predict the price variations for the test dataset. Bayesian Regression estimates for test dataset are computed in the same way as they are computed for train2 dataset – using train1 as an input. 4. Once the price variations are predicted, compute the mean squared error (MSE) for the test dataset (the test dataset has 50 vectors => 50 predictions).

bayesianoptimisation

Code for a Bayesian hyper parameter optimisation algorithm for tuning of ML models

best-readme-template

An awesome README template to jumpstart your projects!