I have tried to investigate the impact of environmental factors and dietary choices on autoimmune flare-ups using a dataset from the Flare Up app. I have aimed to find answers to questions like:
-
Exploratory Data Analysis (EDA): How can we gain a comprehensive understanding of patients data, uncovering hidden insights and patterns?
-
Medication-Symptom Relationships: Is there a discernible connection between the medications patients take and the symptoms they report?
-
Food Categorization: How can we effectively categorize the myriad food items provided by patients, ensuring meaningful groupings through meticulous data cleaning?
-
Correlation with Environmental Factors: Are there correlations between patients environmental conditions (pressure, temperature, humidity) and their self-reported disease activity?
- Understanding the 8 million-row dataset from the Flare Up app.
- Overview of collected data types: symptoms, medications, weather, and dietary information.
-
Merging and cleaning symptom data into meaningful groups.
-
Creating contingency tables to analyze symptom-medication relationships.
-
Investigating correlations between environmental factors (pressure, temperature, humidity) and autoimmune disease activity.
- Employing semi-supervised machine learning, including BERT, to categorize food data as "junk" or "non-junk."
- Exploring correlations between dietary choices and disease activity.







