Hi,
I have been trying to find a thesis topic but i haven't found. I want to to do a thesis in one of these fields Machine Learning , Data Mining, sentiment analysis or Recommender Systems. I have my Thesis topic presentation coming soon but still I don't know what to do. if you have any ideas that i could work on for my thesis please let me know, Thanks in advance.
The whole concept of doing a thesis, and consequently research, is to do something new, building on top of existing science and technology.
Considering the fields you mentioned, Machine Learning , Data Mining, Sentiment Analysis and Recommender Systems, all have a common entity that governs both possibility of success and applicability of research. That entity is data. Now there are two options, i) Collect data ii) Work on existing data. If your plan is to collect data, please assess the time limitations of your thesis deadline since, data collection can be a time consuming process, halting the progress of work. On the other hand if your plan is to work on existing data, there are several places to look into. For example,
- Google Earth Engine - Provides petabytes of georeferenced data i.e. satellite imagery, water level, aerosol level etc. as well as built in machine learning and data mining algorithms.
- National Center for Biotechnology Information - Provides billions of genome sequences to be used for bioinformatics research.
- World Bank Database - Provides economic data on countries across the globe across temporal domain.
The whole concept of doing a thesis, and consequently research, is to do something new, building on top of existing science and technology.
Considering the fields you mentioned, Machine Learning , Data Mining, Sentiment Analysis and Recommender Systems, all have a common entity that governs both possibility of success and applicability of research. That entity is data. Now there are two options, i) Collect data ii) Work on existing data. If your plan is to collect data, please assess the time limitations of your thesis deadline since, data collection can be a time consuming process, halting the progress of work. On the other hand if your plan is to work on existing data, there are several places to look into. For example,
- Google Earth Engine - Provides petabytes of georeferenced data i.e. satellite imagery, water level, aerosol level etc. as well as built in machine learning and data mining algorithms.
- National Center for Biotechnology Information - Provides billions of genome sequences to be used for bioinformatics research.
- World Bank Database - Provides economic data on countries across the globe across temporal domain.
use bibliometrics analysis tools to understand the literature reviews. You can use Histcite, Gephi, Bibexcel , Vosviewer and or R language
You can apply machine learning technology to validate the gene expression signatures that are unique in certain disease, and then use a form of Bayesian modeling algorithm "Hill-Climbing" to identify hidden latent variables involved in networks associated with your disease of choice.
- Badges
- Science topic