mongodb - Is possible to use a Database for storing machine-learning data ? if so, how? -

mongodb - Is possible to use a Database for storing machine-learning data ? if so, how? -

i new machine learning. set machine on server , using database storing info learnt.

var colorclassifier = new neuralnetwork(); colorclassifier.trainbatch([ {input: { r: 0.03, g: 0.7, b: 0.5 }, output: 0}, // black {input: { r: 0.16, g: 0.09, b: 0.2 }, output: 1}, // white {input: { r: 0.5, g: 0.5, b: 1.0 }, output: 1} // white ]); console.log(colorclassifier.classify({ r: 1, g: 0.4, b: 0 })); // 0.99 -

the codes machine learning frameworks this, want 'colorclassifies' logically stored in database , not in memory, can train machine during time without losing old info old trains. don't know how these frameworks internally work, think it's possible asking for. give thanks you

"i don't know how these frameworks internally work"

however framework works, training classifier means finding set weight values such classifier works (usually means minimizing sum of squared errors). so, trained classifier set of real numbers. persist classifier need store these numbers database.

each weight can described 4 numbers:

the layer number (integer): first layer input, rest hidden layers (usually 1 or two), in order appear. from, (integers): since each weight connects 2 nodes, serial number of these nodes within each layer the value of weight (usually real number)

if, example, have weight value 5.8 going 3rd node of 2nd layer 5th node of next layer, can store in table

layer: 2 from_node: 3 to_node: 5 value: 5.8

by repeating weights (a simple loop) can store trained network in simple table. don't know how framework works, there fellow member function returns weights of trained network in list or hashmap format

database mongodb machine-learning artificial-intelligence neural-network