My Weight Loss Journey - The Gradient Descent Way
Few years back, I set a goal to lose 10 kilos. But like most goals, I had no idea how to get there. Should I run daily? Try intermittent fasting? Cut carbs? Count macros? Sleep more? I picked a random starting point - running 5K every alternate day.
A week later, no change. So I limited calories to 1700, down by 0.5 kg. The week after, I cut carbs and added more protein. Down another 1.5 kg. Each week, I observed what worked, adjusted a bit, and moved forward. No perfect plan, just directional corrections.
Years later I realized that I wasn’t just losing weight. I was running Gradient Descent. Not the perfect path, just constant tweaks to get to the result. Seems like Netflix uses gradient descent extensively to minimize prediction error in viewing probabilities.
#DataScience #MachineLearning #DeepLearning
@ShokzUSA Disappointed with the after-sales support from Shokz. My replacement request (Claim No. RMA26050900160100) is being delayed for almost a month because the product is “out of stock” at the service center. Expected a much smoother experience from a premium brand.
Hoping for a faster resolution.
@BlueDart_ Extremely disappointed with the service. For the waybill number 18040829870, the status has been updated as consignee shifted from address which is NOT true. The delivery has never been attempted.
@royalenfield Extremely disappointing experience with Royal Enfield service. Paid ₹11k, but got the bike back with non-functional rear brake, loose clutch, no cleaning, and unresolved issues. Safety compromised. Case ID: CAS-105714293-M1Y4W0.
Service center not helping.