How Predictive Analysis is Predicting Earthquakes, Court Cases and Everything in Between

The predictive analysis model read the transcripts of interviews with 34 at risk youths from New York and 59 at risk youths from Los Angeles and was able to predict which youths would develop psychosis within 2 years accuracy of up to 83%.

No, this isn’t some scene from a science fiction movie, this is a real study from the Mount Sinai school of medicine. The study used predictive analysis, an agile technology to predict with great accuracy which at risk youths would develop psychosis.

What is Predictive Analysis?

Predictive analysis is a series of statistical techniques such as data mining, predictive modeling and machine learning which work together to create models which generate probabilities of future outcomes based on past data.

Why is predictive analysis important?

Predictive analysis is important, because of the current data revolution. Right now businesses are gathering more data than ever, and predictive analysis allows those same businesses to transform that data into solutions and improvements. Predictive analysis also allows managers and business owners to take the difficult business of decision making out of fallible human hands and let the data drive the decision making. This allows businesses to stop reacting and get proactive when it comes to improving everything from supply chains to customer service.

Companies like Workfusion are assisting companies with predictive analysis technologies such as artificial intelligence, cog automation, and crowdsourcing to help implement predictive analysis into your business.

Predictive Analysis works through a 7 step process which looks like this:

  1. Define the project. You first need to define the parameters, data sets and techniques you will be using.
  2. Data Collection. Then you need to collect enough data to have a significant and relevant sample.
  3. Data Analysis. After you collect the data you need to be careful not to overmine, or “torture” it leading to Standard deviations such as those talked about by Professor Gary Smith in his book of the same title.
  4. Statistics. Once you have the cleaned up data you need to validate and test it before creating a model again looking out for biases and errors of omission.
  5. Modeling. Next you will use the validated and tested data to create a model and identify improvements.
  6. Deployment. After the model is validated, you can deploy the improvements through all relevant decisions
  7. Model Monitoring. Lastly you will review the results and make the necessary adjustments.

Now that you know what predictive analysis is and how it works let’s look at 6 new and interesting ways that predictive analysis is being used:

  1. Predicting Mortality. According to Medical News Today , computers can analyze CT images and use them to predict someone’s 5 year mortality rate with 70% accuracy.
  2. Investing. For years investment firms like TrueRisk have used AI and machine learning algorithms to attempt to predict and time markets for investing.
  3. Predicting Psychosis and other mental disorders. As I mentioned earlier, in addition to psychosis, other mental disorders such as bipolar and depression can also be identified by predictive analysis.
  4. Port Fluidity. An unlikely application for predictive analysis, the port of Montreal has deployed predictive analysis to help with better coordination.
  5. Earthquake Prediction. A study published in Geophysical Review Letters has shown that researchers in the UK have collaborated with researchers in the US to create artificial intelligence which can predict earthquakes.
  6. Legal decisions in court cases. Last but certainly not least researchers from Stanford have created a statistical learning model which successfully predicted 70.2% of the Supreme Court decisions from 1816 to 2015.

Predictive analysis or the use of data to create models for predicting the future is no longer science fiction, it’s science fact and if your business wants to keep up it needs to use predictive analysis to gain an edge.

In this article we discussed what Predictive analysis is, why it’s important and looked at the 7-step predictive analysis process which was:

  1. Define project

2. Data collection

3. Data analysis

4. Statistics

5. Modeling

6. Deployment

7. Model Monitoring

And you also learned that predictive analysis was being used to do everything from predict mortality to analyzing supreme court cases.

Now is the time to stop just collecting data and use predictive analysis turn that data into tangible business improvements.