Please enable JavaScript.

Coggle requires JavaScript to display documents.

Machine Learning & Data Access (Data Robot (Upload the file to…

- - - - It is about statistics, knowing what you want the computer to do, what are the right processing steps.
    - - This is the most important aspect of data science, because you have to understand the business problems and data. Thus, in this way you will able to find the solution of the business problem.
    - - You have to be able to write codes and tell the computer what you want to do, know how to hold data sets, build external models, do implementations, everything that involves code activities.
- - - - More sources can produce better improvements on the predictive model.
        
        The more data you have the better decisions you can make . This happens because there is a large gap of acquired information between the company and consumer. By adding more data it can reduce this gap, so in this way the firm can make better decisions when it is about providing value to the consumer.
        
        Also allows to make more decisions on more people. There is a lot of people when the company has few sources to make predictions. It is not enough. By having more data, these decisions can be more accurate for broad potential customers.
  - - - Data is growing everyday, it is difficult to handle large amount of data because companies can loose important opportunities in reaching potential consumers.
- - - - DemystData:
        
        Starts with model creation where the demystdata have you rapidly test data, discover new data, combine with more data, build new features, etc.
        
        DenymstData provides a lot of opportunities to achieve competitive advantage, there are tow keys to become more effective in achieving this goal:
        
        (1) Teach everyone in the company how to identify opportunities by using machine learning. This, they can make positive impacts to the company.
        
        (2) Automation and not prioritization
        
        Real-Time Data Access
        
        DemystData has a platform that connects data sources such as traditional, non-traditional, online, Telco, etc. The problem is that data is unstructured, there is a schema that is more usable and helps the user to understand. By applying the logic and structured it makes easier to use different sorts of data, they also applied standardization, they use different keys to match together.
        
        Left Graph --> represents the unstructured data.
        
        Right Graph --> represents the DemystData platform or schema.
      - DataRobot
        
        Allows you run through these features and find the importance and find how to combine them to build more powerful factors.
- - - - Many manual tasks are highly automatable, that makes execution to be dramatically accelerated.
        
        Example
        
        DemystData Platform
        
        Takes data from different sources, finding out signals from within those sources, combines them, and use those sources in action. Usually to complete this process takes a year when you use DemystData Platform.
        
        Components:
        
        Data = it can be traditional, non-traditional, web search, etc.
        
        Connectors = connects to a data source
        
        1 more item...
        
        Process
        
        1 more item...
        
        Data Scientist
        
        Data scientist takes a day, they use individual attributes from existing and new sources. It combines them in new ways, and finds packet of value:
        
        Customers declining and has lower risks comparing with most of our customers (They are represented by the green circles from the graph).
        
        Negative customers represented by the red circles, they seem to be good. However, when we use additional data it shows that these customers ended up being highly fraudulent.