Is Predictive Analysis Reliable?

“Predictive analytics is the branch of the advanced analytics which is used to make predictions about unknown future events. Predictive analytics uses many techniques from data mining, statistics, modelling, machine learning, and artificial intelligence to analyze current data to make predictions about the future.” - www.predictiveanalyticstoday.com/what-is-predictive-analytics/

The philosophy of Machine Learning and Artificial intelligence are based on the belief that they should be able to sense, reason, act, and then adapt based on experience. A brief description of each is given below

Sense - This involves Identifying and recognizing meaningful objects or concepts in the midst of vast data.

Reason -Understanding the larger context. 

Act - Recommending the appropriate course of action.  

Adapt- The ultimate stage is that we must be able adapt algorithms at each phase based on experience, retraining them to be ever more intelligent.  

The challenge for Predictive Analysis is that it is based on the premise of accurately identifying the right data parameters. If you get the wrong set of data, or there is a flaw in even one set the whole premise may come tumbling down.

A very important myth of Predictive Analysis tools was shattered when results for the recently concluded elections in the US were announced on 8th November.  During the entire course of elections a perspective had been floated based on Predictive Analysis that a particular candidate would win. However the results were a surprise to many pundits when the accuracy of the prediction proved wrong.

Somewhere some important data was missed or overlooked and all predictions were found to be wanting.

In 1952, Ray Bradbury wrote a very interesting science fiction story titled "A Sound of Thunder". The basic premise of the story is that in a future date, 2055 time travel has become possible. It is also a period when there are two candidates in the presidential elections, a fascist candidate, Deutscher, and a moderate candidate Keith.  The story goes on to state that one traveller goes back to the dinosaur era, just after the announcement of the results, wherein the moderate candidate Keith had won. During the trip the traveller inadvertently kills a butterfly and on his return finds that the new president is Deutscher and not Keith. Changing the whole sequence of events.

One small event in the past had a major impact on the future.

Coming back to the election results of 8th November, where one candidate was probably led into a state of complacency based on Predictive Analysis of her win, can we assume that “Predictive Analysis” is the future for decision making? Or is it still work in progress?  And still needs more maturity? Can one small change in data bring about such a major difference in the results. (Disclaimer - This is not to say either candidate is a moderate or fascist)