Alex Caselli's repositories
Federated-Learning-for-Human-Mobility-Models
Thanks to the proliferation of smart devices, such as smartphones and wearables, which are equipped with computation, communication and sensing capabilities, a plethora of new location-based services and applications are available for the users at any time and everywhere. Understanding human mobility has gain importance to offer better services able to provide valuable products to the user whenever it's required. The ability to predict when and where individuals will go next allows enabling smart recommendation systems or a better organization of resources such as public transport vehicles or taxis. Network providers can predict future activities of individuals and groups to optimize network handovers, while transport systems can provide more vehicles or lines where required, reducing waiting time and discomfort to their clients. The representation of the movements of individuals or groups of mobile entities are called human mobility models. Such models replicate real human mobility characteristics, enabling to simulate movements of different individuals and infer their future whereabouts. The development of these models requires to collect in a centralized location, as a server, the information related to the users' locations. Such data represents sensitive information, and the collection of those threatens the privacy of the users involved. The recent introduction of federated learning, a privacy-preserving approach to build machine and deep learning models, represents a promising technique to solve the privacy issue. Federated learning allows mobile devices to contribute with their private data to the model creation without sharing them with a centralized server. In this thesis, we investigate the application of the federated learning paradigm to the field of human mobility modelling. Using three different mobility datasets, we first designed and developed a robust human mobility model by investigating different classes of neural networks and the influence of demographic data over models' performance. Second, we applied federated learning to create a human mobility model based on deep learning which does not require the collection of users' mobility traces, achieving promising results on two different datasets. Users' data remains so distributed over the big number of devices which have generated them, while the model is shared and trained among the server and the devices. Furthermore, the developed federated model has been the subject of different analyses including: the effects of sparse availability of the clients; The communication costs required by federated settings; The application of transfer-learning techniques and model refinement through federated learning and, lastly, the influence of differential privacy on the model’s prediction performance, also called utility
Evolutionary-Strategies-Benchmark
The scope of this project is to implement and test three different evolutionary strategies (Cross-Entropy Method (CEM), Natural Evolution Strategy (NES), Covariance Matrix Adaptation Evolution Strategy (CMA-ES)) on two different convex functions (a sphere function and a 2-dimensional Rastrigin function) to further explore their capabilities.
NodeJs-Unsplash-Proxy
Simple NodeJs Proxy running on Google App Engine to interact with Unsplash API using you API Key
PERSONa-Mobility-Dataset-Generator
This dataset born from the need of mobility traces provided with demographics data of the users and it allows to define several classes of users with their most relevant places. Using probability distributions, it can be used to generate slotted mobility traces for different users.
Deep-Q-Network-Atari-Breakout
The purpose of this project is to train a Deep Q-Network agent (https://daiwk.github.io/assets/dqn.pdf) using the OpenAI Gym environment (https://gym.openai.com/) to play the famous Atari game BreakOut. The DQN agent has 3 main components: the online Q-network, the target Q-network, and a replay buffer.
House-Activity-Prediction-with-HMM
In this project we had to inference about user activities in a house relying on registred action by different sensors. After model building we created a demo, for demo running it's necessary launch dash_house.py in Code folder. The report is in Italian.
Chess-Recognition-MatLab
MatLab application to read printed chessboards on photos
MMORPG-Game-Data-Analytics
In this project we had to analyze at MMORPG game network in a certain period (from 12/1/2009 to 12/31/2009); the nodes are users and the edges are the type of relationship between them (trade, message and attack). The report is in Italian, the demo was done with Dash (a python library); for running demo go to dash folder and run project_data.py.
Text-Generation-Using-RNN
Give a large book from Project Gutenberg (e.g., The Count of Monte Cristo). Train a network to predict network trained to predict the next textual character given a sequence of characters. Such network can be used to generate text by sampling a character
Usability-Analysis-of-Fitness-Apps
In this academic project, we made a usability study of two of the most popular fitness apps used to follow a diet.