Ricardo Arias Salazar's Projects
Wrangling a dataset that contains transactional retail data from an online electronics store (DigiCO) in Melbourne, Australia.
Accuracy comparison between generative and discriminative models
Design of a model with appropriate feature engineering, that estimates one target temperature rotor temperature (βpmβ) in a causal manner, based on the data set that records the rotor temperatures of a permanent magnet synchronous motor (PMSM) in real-time
Classifying abstracts of different papers using unsupervised learning algorithms like soft and hard Expectation Maximization.
In this project, we use Spark to visualize, manipulate, model and stream historical flight-delays data using Spark RDD, Spark SQL and Kafka
Integration of 7 different datasets in various formats about housing information in Victoria, Australia. And study the effect of different normalization/transformation methods
The aim of this project is to explore three different datasets related to the Indigenous population in Australia, and get some insights about the region, age, and immunisation rates
The inventory system simulates the stocking level and the revenue of Cantilever Umbrellas in an Australian firm.
Function which takes training data and their labels, and the size of the neighborhood (K), and then it returns the regressed values for the test data points.
Using a perceptron to classify three different classes in a dataset.
Simulation of social links and disease transmission between people, over a period of time, and plot graphs to determine whether an outbreak is contained or not
Using three different datasets, obtained from three different places we built a dashboard that could be helpful for different people to identify what is the relationship between poverty (measured by the Multidimensional Poverty Index), communes, house pricing.
This is an interactive map that integrates the location of the top recreation sites around Victoria, Australia with the Air Quality index in each suburb. Allowing people to choose a perfect recreation spot based on the quality of the air
Use the different techniques of resampling to quantify the uncertainty of predictions for a KNN regressor
Load monitoring/ load detection is one big breakthrough in tackling the problem of increasing carbon footprint. It helps to provide detailed electricity consumption information in residential households. This project is dedicated to providing a perfect estimate of the usage of the most common appliances in residential buildings.
π
There are three classes InfoTheory, CompVis and Math. These can occur in any combination, so an article could be all three at once, two, one or none. The job is to build text classifiers that predict each of these three classes individually using the Abstract field.
Extraction of data from semi-structured text files, and preprocess the text into numerical representations.
Dashboard using the data of the trees located in Fitzroy Garden in Melbourne, Australia
Extract data from semi-structured text files and transform the data into XML format.