Where could I learn more about ML for this specific problem

Hi All,

Hopefully someone will be able to help me out here, I'm a full stack dev looking to venture in to ML to solve a specific problem. I'm looking for the best reading materials for where to start with this problem and get in to ML. I apologize if this is too generic, I'm trying to get a good place to start that applies to the problem I'm trying to solve.

The problem; predict the outcome of an event (either a +1 or -1 result) based on a time series of data properties for an object.

I have a dataset that contains a time series. Each time series entry would be structured as {recordId: id, timestamp: eventDate, prop1: value, prop2: value, prop3: value} and so on. For each time series one or more of the properties has changed for a given id. So I can trace the history of this object from inception until +1 or -1 (also a property on the object). I would like to train based on the historical data I have, and then be able to pass in a current object and have it output the probability of a +1 or -1 outcome based on the current properties and the history of the object.

Example data: {recordId: 1, eventDate: 1/1/2013, status: 'new', source: 'web', result: 0} {recordId: 1, eventDate: 1/2/2013, status: 'in progress', source: 'web', result: 0} {recordId: 1, eventDate: 1/3/2013, status: 'resolving', source: 'web', result: +1}

{recordId: 2, eventDate: 1/1/2013, status: 'new', source: 'phone', result: 0} {recordId: 2, eventDate: 1/4/2013, status: 'in progress', source: 'phone', result: 0} {recordId: 2, eventDate: 1/9/2013, status: 'resolving', source: 'phone', result: -1}

Then I would like to input the following: {recordId: 3, eventDate: 1/1/2013, status: 'new', source: 'web', result: 0} and have it output the odds this results in a +1 result.

Any help on where to start, books to read, would be super!

Thanks!