Exploratory Data Analysis (EDA) and Visualization Techniques..
summary
Checks whether masks actually cover useful pixels not just counts uses pixel spacing to convert mask pixels to physical area and finds 12k chest x‑rays with heavy class imbalance but measurable size differences across age and sex uses a pre‑trained lung mask that works on many male images but fails on many female ones counts defect instances with OpenCV contour counting to show class1 is many tiny spots while class4 is few large patches uses frequent‑pattern mining on 5.5k cloud images to expose common co‑occurrences and guide augmentation and training decisions.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 3:38Image segmentation challenges
- 8:15Pneumothorax data analysis
- 11:04Steel defect detection imbalance
- 16:38Defect types and sizes analysis
- 18:24Cloud classification insights
- 21:57Key takeaways from EDA and visualization techniques
- 23:04Closing
related talks
37mUsing Exploratory Data Analysis to Discover Patterns
27mCelebrating Neurodiversity Through Aesthetic Data Visualization
01h 00mData, visualization, and designing AI
32mEDA Techniques for Data Analysis
38mTransforming Data to Unlock Its Latent Value
40mVisual Analysis for Data Science
22mCOVID-19: Data Science & Expertise
35mDeep Learning Model Interpretability for Computer Vision based Models