Have a look at Kafka Connect (http://docs.confluent.io/2.0.0/connect). The JDBC Connector will poll for database changes changes and push them to a Kafka topic. Means you should see all the changes in the database rather than a snapshot say once a day
Thanks for sharing your experience! Couple of questions
Why implement the training in NodeJS and not use an existing library in R or Python (scikit-learn) and just implement the scoring (feedforward network) in Node?
Did you just use a single test/train split? What is the variation in Res if you run cross validation?
Your article suggests that you used MI to select the 10k best features. Did you perform this MI feature selection before your test/train split? If so, you would already be "using" your class labels, and the results will be biased. It is likely your true generalisation error will be lower.