Via Hacker News:

Ariel lost 20 lbs of weight and documented his progress using a series of scripts. He utilized the Vowpal Wabbit (VW) machine learning project. The VW project is co-sponsored by Microsoft and Yahoo Research. Ariel had logged his data as a simple CSV file using only weight, sleep, exercise and foods consumed as data points. He then wrote a perl script to convert his CSV over to the vowpal-wabbit traning-set.

He learned which foods, exercise and sleep patterns showed a positive correlation with weight loss. The takeaways from his N=1 study are:

• Sleeping longer consistently appeared as the #1 factor in losing weight
• Lack of sleep did the opposite: too little sleep lead to weight gains.
• Lack of sleep did the opposite: too little sleep lead to weight gains
• Carbs made me gain weight. The worst were high-starch and sugary foods
• Fatty and oily foods did the opposite: they were positively correlated with weight-loss
• Skipping breakfast and Stop eating earlier in the evening before going to bed.