Showing posts from October, 2016

Good Health numbers

HDL (good cholestrol) > 50 LDL (bad cholestrol) < 100 (best is < 50 - that is centenarian type)
Total cholestrol (HDL + LDL) < 200
High cholestrol means heart attack risk (#1 killer) Obesity, diabetes, high blood pressure and smoking are 4 other risk factors.
CHOLESTEROL ------------------ Cholesterol ---<200 HDL---40---60 LDL---<100 VLDL --<30 Triglycerides --<150 ----------------------------
CHOLESTEROL ---------------- Borderline --200 -- 239 High ---->240 V.High -->250 ----------------------------
LDL ------ Borderline --130 ---159 High ---160---189 V.High --> 190 ----------------------------
TRIGLYCERIDES ----------------- Borderline - 150 -- 199 High --200---499 V.High -->500 ----------------------------
PLATELETS COUNT ---------------------- 1.50Lac----4.50 Lac ----------------------------
BLOOD ----------- Vitamin-D --50----80 Uric Acid --3.50---7.20 ----------------------------
KIDNEY ---------- Urea---17---43 Calcium --8.80--10.60 Sodium --136---146 Pr…

A Virtual assistant for Raspberrry Pi

It is made following the nice book on the subject by Tanay Pant - Building a Virtual Assistant for Raspberry Pi (APress) This is similar though much simplified version of products like Google Home or Amazon Echo. The idea is to use the Raspberry Pi based Virtual Assistant as one of the interfaces for Amigo chatbot project. Ria uses Google STT (Speech to text) API and espeak on Linux (or say on OSX) for TTS (Text to speech). Currently supported skills/features: Play music from configured path on local diskRead news headlines scrapping news from notes and store in local sqllite DBTell timeTell local weather of configured cityPost tweets on twitterLook up wikipedia and read a short description for any term, person or place (anything on wikipedia).
Source available on Github:

Movie Recommender System (CMPE 277)

Movie recommender built using Apache Spark + mongo DB + Android App

Uses Spark MLlib model based collaborative filtering for movie recommendation (Alternating Least Squares algorithm).

Project sources are at: - Apache Spark + Mongo DB based backend service for movie recommendations - Android App