MindsDB 0.6.7

MindsDB's goal is to make it very simple for developers to use the power of artificial neural networks in their projects.

MindsDB

MindsDB's goal is to make it very simple for developers to use the power of artificial neural networks in their projects.

Getting started

It's very simple to setup

 pip3 install mindsdb --user

To train a model:

from mindsdb import MindsDB

mdb = MindsDB()
mdb.learn(
    from_file='monthly_sales.csv',
    predict='sales',
    model_name='sales_model'
)

To use a Model:

from mindsdb import MindsDB

mdb = MindsDB()
predicted_sales = mdb.predict(predict='sales', when={'month': 'Sept'}, model_name='sales_model')

Note: You must have a MongoDB server running, by default, MindsDB will look for it in mongodb://127.0.0.1/mindsdb

Note2: MindsDB uses sqlite as its sql engine, it is recommended that you update to have the latest sql in your system we suggest versions >=(3.25.0):

On mac:

brew upgrade sqlite

CONFIG Settings

You can manipulate the default settings of MindsDB by setting env variables before importing the mindsdb module

For example: if you want to specify a different Mongo Server

import os

os.environ['MONGO_SERVER_HOST'] = 'mongodb://[username:password@]host1[:port1][,host2[:port2],...[,hostN[:portN]]][/[database][?options]]'

# now we import mindsdb
from mindsdb import MindsDB

Why use MindsDB?

Developers today are more aware of the capabilities of Machine Learning, however from ideas of using ML to actual implementations, there are many hurdles and therefore most ideas of using Machine Learning never even start.

Thanks to MindsDB people building products can focus more on:

• Understanding what problems/predictions are interesting for the business.
• What data should be of interest for a given prediction.

Less on: spending countless hours building models, making data fit into such models, trainining, testing, validating, tunning hyperparameters, ....