Hi, my name is André Ip
And I'm a full time learner!

The Code It challenge was organized by NEEC "Núcleo de Engenharia Eletrotécnica e Computadores" from NOVA University of Lisbon with the partnership between KPMG and Outsystems. The challenge consisted to solve an issue selected by us. We decided to implement a search engine for Social Events. Within seven days we manage to achieve a full functional app which provided us Events information from Lisbon.

In this project were elaborated a streaming and remote system (raspberry pi, servo, camera) controlled by other device (computer). Was used a servo with an attached camera. Furthermore, was implemented a class in Python to control the servo movements. After this part of the work I was able to start the communication implementation between the system and the device. HTML and JavaScript were the fundamental keys for this step where was created the streaming system. In the end of the project we could control the servo with the camera on a local network (server-client) from the device, in this case was used a computer running in linux.

My internship with EDP was related with Data Analysis. Firstly, the data was retrieved from the SAP platform, thereafter were conducted querie arrangements in order to gather the best possible data visualizations. The Microsoft Power BI tool was selected to visualize the data related with the Energy incidents from the south of Portugal. In the half way through the internship I could gather enough information of the low voltage division and several data insights to make a presentation to the department director, my engineer adviser and for work colleagues (😄 Picture ➯). The following weeks I implemented data arragements and visualizations for the medium and high voltage divisions. The data visualizations were also implemented online and mobile.

The main goal of my Master Thesis was the construction of a Geolocation Prediction Architecture (GPA) on top of a mobility data repository from the city of Porto. The repository contains a large dataset referring taxis mobile movements. Thereafter, we performed several pre-processing insights for the mobility data. Consequently, we built two approaches for estimation and prediction of future positions from an mobile entity. The first approach involves probabilistic methodologies with Hidden Markov Model. The second approach explores deep learning with a Recurrent Neural Network. During the process of the thesis were performed data analysis to gather the greatest results for the predictions accuracies.