marcroiglama Goto Github PK

911_calls_case_analysis

A small data analysis of USA emergency call register. An example of cleaning and transforming data to get a better understanding of the data. Moreover, some findings will be shown, take a look at the Jupyter notebook for some data fun!

a-b-test-on-search-system-second-hand-app

bike_sharing_analysis

The data is coming from https://citybik.es/, a repository of all the bike-sharing sevices around the world. Using the script tempo.py it's possible to get the real-time data from the web and work with it. The code must be changed indicating the service you want to analyze, the example is for Bicing from Barcelona.

computer_vision

Computer Vision Porfolio.

credit_fault_detection

dbconnectors

featureanalyzer

givo

GIVO (Garbage In, Valuable data Out) is a entire system to enhance the quality of your data using unsupervised Machine Learning methods. Takes advantatge of three outlier detector methods (Isolation Forest, Elliptic Envelope and Local Outlier Factor) to create an unique Ensemble Outlier Detector to automatically detect wrong mesures in environmental sensors data.

machine-learning

IThe aim of this work is to catalog several machine learning algorithms implementations, test their performance on real data and play with the hyperparameters.

nlp-product-classifier

pandas_engineering

When we are working large files with pandas library we can suffer from memory errors or slow processing as Pandas is a very powerful tool but very memory consuming in terms of RAM. On this git I present a simple way to reduce the memory overload of pandas dataframes using pandas formatting and some transformations.

sensor_calibration

Machine Learning supervided models for sensor signal calibration on environmental data.

streaming-platform-dataset-analysis

wines_reviews

The data is coming from https://www.kaggle.com/zynicide/wine-reviews and it's a massive dataset of wines reviews with information as the professional description, price, points and so on. The aim of the project is to give an overview of a worldwide wine clues, for example, the relation between price and quality(points), the countries where the wines are made or perform an analisys of the most used words to decrive wines. The future lines of this analisys could be to implement a Machine Learning algorithm to predict the clues(aroma,taste) that corresponds to an exellent wine.