Neural Networks in the Mining Industry
A significant issue in the mining industry involves the amount of accidents and casualties caused by falling rock. In the past, miners had to test whether or not an area was safe to work in by striking the hanging wall with a sounding bar and listening for the sound it made. While many skilled miners could hear the distinction, it was still not an extremely accurate practice. Thus, the Council for Scientific and Industrial Research decided to design a device that will help miners conclude whether or not an area is safe to work in on an objective basis. The technology that is used in this new tool is an artificial neural network that records the noise from the sounding bar, processes it through the layers of the model, and then categorizes the wall as either “intact or detached”.1 The device has proven accurate more than 70% of the time and could potentially save many lives in the future.
Neural Networks in Marketing
Microsoft has put artificial neural networking to the test in their marketing promotions. Microsoft mails out advertisements to customers annually to remind them to upgrade their systems or to entice them to buy new items. In the past, Microsoft could expect about 4.9% of the people who received something in the mail from them to respond. However, they have found a way to increase this percentage by using a neural networking software called BrainMaker to determine which customers were the most likely to respond. BrainMaker was used to figure out what attributes of a customer made them more likely to respond to these ads. Ultimately, Microsoft was able to increase their percentage of response from 4.9% to 8.2%.2
Linear Programming in Healthcare
Linear programming is used in the healthcare industry for reducing wait times in hospitals and doctors’ offices. This is both beneficial in terms of convenience and in terms of safety. Often times, long waiting periods can be unfavorable to a patient’s health and thus, shortening these wait times could potentially save lives. Researchers in Canada have found that using a combination of two scheduling models, the Markov Decision Process model (MDP) and the Approximate Dynamic Programming (ADP) method, can help with scheduling appointments based on the severity of the patient’s conditions.3 This will make optimal use of the healthcare providers’ time and cut down on waiting times for the severely ill.
Scheduling and Optimization
Scheduling and optimizing time is a challenge for many businesses, but linear programming can help. The Fred Astaire East Side Dance Studio in New York needed an efficient way to schedule dancers in its semiannual showcases. They used a combination of Visual Basic and Excel to create a model that could create an optimal schedule of the performances for the showcases. The program scheduled 583 heat entries, 19 different types of dance, 18 solo dances, 28 students, and eight teachers in the 2007 summer showcase.1 Not only did this help the showcases run smoothly, but it also drastically decreased the amount of time that workers had to spend making the schedule.
References
1. Sharda, R., Delen, D., Turban, E. (2014). Business Intelligence and Analytics: Systems for Decision Support 10th ed. 2. Holcombe, Colin. Neural Network Analysis. Retrieved from http://www.ecommerce-digest.com/neural-networks.html 3. Patrick, Jonathan. Puterman, Martin. (February 2008). Reducing Wait Times Through Operations Research: Optimizing the Use of Surge Capacity. Retrieved from http://lawandgovernance.com/content/19575
Opmerkingen