Computer Technology How machines learn

Feb. 10, 2015 - PRLog -- Machines function in algorithms. The word algorithm sounds intimidating at first, but when defined an algorithm is not more than a set of patterns a machine acts upon as it performs the assigned task.

The two most common types of algorithms are

A. statistical reasoning
B. inductive reasoning

In statistical reasoning the machine simply gathers data and analyzes the probabilities of future occurrences following prior results.

For example, 90% of crows are black then it is highly probable that 90% of all crows will be black.
When talking about inductive reasoning, we are looking at analyzing probabilities. These probabilities are geared toward proving or disproving a theory.

These probabilities are based upon gathering instances of events. For example, startled dogs will snap when they are surprised.

Random results will not be included until they represent a significant percentage.
Search engine results are based upon the algorithms we have discussed. Search engines can’t measure a user’s emotions, but they can measure the amount of time a person spends on a website. For example most search engines consider a typical visit to a page lasting under 30 seconds to be a negative experience, which is based upon past algorithms indicating a user didn’t find what they were looking for. Search engines know this because they follow a user as they move from site to site until a solution is found.

In inductive reasoning, search engines look at many factors including:

1. bounce rate
2. time on page
3. actions users take when on that page.

Inductive reasoning uses a series of signals including the time signal we discussed above.
Sentiment analysis

Sentiment over the Internet plays a big part in search ranking. Reviews are important. Search engines focus upon words including, excellent, fantastic, great, bad, poor and horrible. A computer program or spider can’t read a webpage like a human, but it can look for word patterns. Would you like to find companies the Internet considers to be bad examples of customer service? Simply type bad customer service companies into Google and Bing then analyze the results. The Internet will reveal sites that will provide names of companies who are accused of delivering bad customer services. For this reason, pay close attention to your online reputation.

Repeat the same experiment by typing poor customer service companies and analyzing the company names.

If you want to learn more about the operations of spiders and search when it comes to ranking your webpage, contact ABCO Technology Monday through Friday from 9 AM to 6 PM. Call us at (310) 216-3067.

For more information visit us online.

https://www.pinterest.com/abcotechnology/abco-technology-...

http://www.yellowpages.com/los-angeles-ca/mip/abco-techno...