Innovative Research Projects in Signal Processing, Tourism, and Language Translation
1
st
EURASIP-GAIPDM Seasonal School on
“Learning from Signals, Images, and Video”
HACKATHON-I
Organizers:
C. Kotropoulos
M. Ntemi
G. Karantaidis
I. Sarridis
E. Gionanidis
Tourism
Recommendation
Problem formulation:
•
Recommend Places of Interest (POIs) using adaptive hypergraph
learning.
Input:
•
POIs, which are represented by geo-clusters.
Output:
•
optimal ranking vector
f
. The sorted
f
contains the recommended POIs in
descending order.
•
Recall-Precision curve.
Research ideas:
•
Fine tuning of weight matrix
W
parameters.
•
Fine tuning of incidence matrix
H
parameters.
•
Combine
W
and
H
matrix updates.
•
More ideas and detailed descriptions can be found in
README.txt
file.
Price
Prediction
•
Problem: Given the past airfare prices predict the next
price
•
Dataset: 27 routes.
–
1 route departure from New York, 26 from European cities
–
14 arrival to Athens, Greece, 13 arrival to Thessaloniki,
Greece
•
Input: " days_until_flight, departure, arrival,
ticket_price "
•
Output: airfare price prediction
•
Approaches: Collaborative Kalman filter, Particle filter,
LSTM NN, etc.
Hotel Rating Prediction
•
Problem: anticipate the ratings that will be
given to hotels from users with similar taste
•
Dataset: 196,536 ratings from 170,187 users
and 1,097 hotels in Santorini, Greece
•
Input: "hotel_id, user_id, rating, date“
•
Output: Hotel rating
•
Approaches: Matrix factorization techniques,
Collaborative Kalman filter, Collaborative Deep
Learning etc.
Natural Language
Processing
Problem:
given a sentence (or text) in one language, find the
translated one in another language.
Dataset:
•
pairs of sentences (192,881) in English and in
German
•
e
.g., I am safe / ich bin sicher
Input:
the text in the input language
Output:
the translated text
Approaches:
Statistical Machine Translation, Neural Machine
Learning, Rule-based and corpus-based machine translations
This text introduces several cutting-edge research projects in different domains, including signal processing, tourism recommendation systems, price prediction, hotel rating anticipation, and natural language processing for translations. Each project aims to solve unique challenges using advanced techniques in machine learning and data analysis, showcasing the diverse applications of these fields.
Download Presentation
Please find below an Image/Link to download the presentation.
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author. Download presentation by click this link. If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.
E N D
Presentation Transcript
1stEURASIP-GAIPDM Seasonal School on Learning from Signals, Images, and Video HACKATHON-I Organizers: C. Kotropoulos M. Ntemi G. Karantaidis I. Sarridis E. Gionanidis
Tourism Recommendation Problem formulation: Recommend Places of Interest (POIs) using adaptive hypergraph learning. Input: POIs, which are represented by geo-clusters. Output: optimal ranking vector f. The sorted f contains the recommended POIs in descending order. Recall-Precision curve. Research ideas: Fine tuning of weight matrix W parameters. Fine tuning of incidence matrix H parameters. Combine W and H matrix updates. More ideas and detailed descriptions can be found in README.txt file.
PricePrediction Problem: Given the past airfare prices predict the next price Dataset: 27 routes. 1 route departure from New York, 26 from European cities 14 arrival to Athens, Greece, 13 arrival to Thessaloniki, Greece Input: " days_until_flight, departure, arrival, ticket_price " Output: airfare price prediction Approaches: Collaborative Kalman filter, Particle filter, LSTM NN, etc.
Hotel Rating Prediction Problem: anticipate the ratings that will be given to hotels from users with similar taste Dataset: 196,536 ratings from 170,187 users and 1,097 hotels in Santorini, Greece Input: "hotel_id, user_id, rating, date Output: Hotel rating Approaches: Matrix factorization techniques, Collaborative Kalman filter, Collaborative Deep Learning etc.
Natural Language Processing Problem: given a sentence (or text) in one language, find the translated one in another language. Dataset: pairs of sentences (192,881) in English and in German e.g., I am safe / ich bin sicher Input: the text in the input language Output: the translated text Approaches: Statistical Machine Translation, Neural Machine Learning, Rule-based and corpus-based machine translations
