marcroiglama
🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍 🐍
DelectatechBarcelona
Pinned Repositories
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.
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.
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.
marcroiglama's Repositories
marcroiglama/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.
marcroiglama/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.
marcroiglama/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.
marcroiglama/featureAnalyzer
marcroiglama/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.
marcroiglama/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!
marcroiglama/A-B-test-on-Search-System-Second-Hand-APP
marcroiglama/Computer_vision
Computer Vision Porfolio.
marcroiglama/credit_fault_detection
marcroiglama/dbconnectors
marcroiglama/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.
marcroiglama/NLP-product-classifier
marcroiglama/Sensor_calibration
Machine Learning supervided models for sensor signal calibration on environmental data.
marcroiglama/Streaming-Platform-Dataset-Analysis