Artificial Intelligence on Business Forecasting
The business world today is driven by customer demand. Unfortunately, the demand
patterns vary considerably from time to time. That is why it can be so challenging to
develop accurate forecasts. The goals of forecasting are to reduce uncertainty & provide
benchmarks for monitoring actual performance.
A new generation of Artificial Intelligence technologies have emerged that hold significant commitment on improving forecasting process including applications as product demand,
employee turnover, cash flow, distribution requirements, manpower forecasting & inventory.
Artificial Intelligence systems are designed to bridge the gap between the two traditional
forecasting approaches - Managerial and Quantitative.
Artificial Intelligence is defined as a Computer Based analytical process exhibiting actions
and behavior considered ' Intelligent' by human observers. A.I. attempts to mimic human
thought process including reasoning and optimization. One purpose of A.I. is to help
organize and supply information for Management decision making in such a way as to
improve overall efficiency and performance.Three of the more commonly used A.I.systems
in forecasting are : -
1) Expert Systems - Summarize totality of available knowledge & rules. "Knowledge" is
stored in a set of "if-then" rules. Knowledge base can be made by interviewing experts or
integrating sets of Data. For example, predicting upcoming weather conditions based on
current temperature. Humidity levels, season of year and Geographical location.
2) Neural nets - Emulate elements of the Human cognitive process, especially the ability
to recognize and learn patterns. The design consists of a large number of nodes that serve
as calculators to process inputs & pass the results to other nodes in the network. These systems
have the advantage of not requiring prior assumptions above possible relationships. One
application of the neural nets might be forecasting employee turnover by category based on
such factors as tenure with the company, Gender and the Managerial level.
3) Belief Networks - It describe Data Base structure using a Tree format. Nodes represents
variables and Branches the conditional dependencies between variables. Beliefs net generate
conditional probabilities for a number of future outcomes. For example, estimating chances
of product sales levels based on such traditional factors e.g. marketing, R & D budget as also
market signals like customers' complaints.
These systems can be employed for both forecast classification e.g. Preferred customer Vs.
Marginal customer and Prediction e.g. Annual Sales. It can be used to refine marketing strategy
based on three customer behavior factors - Profit Margin, Retention Probability and Potential
long term gain to the company. It is a matter of concern to note that big organisations have
improved Bottom line profitability by implementing the forecasting process. Few examples
to glance into the usage benefits like - Hyundai Motors (reducing delivery time by 20%) and increased Inventory turns from 3 to 3.4,Unilever reduced forecasting error from 40% to 25%
yielding results in multi-million dollars savings.In a competitive Business environment,
forecasting can lead to significant advantages as well as to costly mistakes.
Following process outlines a plan for improving forecast accuracy using Artificial Intelligence
support systems.
1) Evaluate & Characterize the current forecasting system.
2) Measure current level of error.
3) Compare error levels with Industry norms.
4) Specify new requirements.
5) Identify alternative forecasting options.
6) Characterize economic impact of improved forecasts,
7) Select best approaches.
8) Develop implementation schedule.
9) Identify potential problems and bottle neck areas.
10) Implement new system & monitor performance.
Primary object of using Artificial Intelligence is to better integrate Managerial and Quantitative
estimates thereby reducing forecasting errors.
- - -
patterns vary considerably from time to time. That is why it can be so challenging to
develop accurate forecasts. The goals of forecasting are to reduce uncertainty & provide
benchmarks for monitoring actual performance.
A new generation of Artificial Intelligence technologies have emerged that hold significant commitment on improving forecasting process including applications as product demand,
employee turnover, cash flow, distribution requirements, manpower forecasting & inventory.
Artificial Intelligence systems are designed to bridge the gap between the two traditional
forecasting approaches - Managerial and Quantitative.
Artificial Intelligence is defined as a Computer Based analytical process exhibiting actions
and behavior considered ' Intelligent' by human observers. A.I. attempts to mimic human
thought process including reasoning and optimization. One purpose of A.I. is to help
organize and supply information for Management decision making in such a way as to
improve overall efficiency and performance.Three of the more commonly used A.I.systems
in forecasting are : -
1) Expert Systems - Summarize totality of available knowledge & rules. "Knowledge" is
stored in a set of "if-then" rules. Knowledge base can be made by interviewing experts or
integrating sets of Data. For example, predicting upcoming weather conditions based on
current temperature. Humidity levels, season of year and Geographical location.
2) Neural nets - Emulate elements of the Human cognitive process, especially the ability
to recognize and learn patterns. The design consists of a large number of nodes that serve
as calculators to process inputs & pass the results to other nodes in the network. These systems
have the advantage of not requiring prior assumptions above possible relationships. One
application of the neural nets might be forecasting employee turnover by category based on
such factors as tenure with the company, Gender and the Managerial level.
3) Belief Networks - It describe Data Base structure using a Tree format. Nodes represents
variables and Branches the conditional dependencies between variables. Beliefs net generate
conditional probabilities for a number of future outcomes. For example, estimating chances
of product sales levels based on such traditional factors e.g. marketing, R & D budget as also
market signals like customers' complaints.
These systems can be employed for both forecast classification e.g. Preferred customer Vs.
Marginal customer and Prediction e.g. Annual Sales. It can be used to refine marketing strategy
based on three customer behavior factors - Profit Margin, Retention Probability and Potential
long term gain to the company. It is a matter of concern to note that big organisations have
improved Bottom line profitability by implementing the forecasting process. Few examples
to glance into the usage benefits like - Hyundai Motors (reducing delivery time by 20%) and increased Inventory turns from 3 to 3.4,Unilever reduced forecasting error from 40% to 25%
yielding results in multi-million dollars savings.In a competitive Business environment,
forecasting can lead to significant advantages as well as to costly mistakes.
Following process outlines a plan for improving forecast accuracy using Artificial Intelligence
support systems.
1) Evaluate & Characterize the current forecasting system.
2) Measure current level of error.
3) Compare error levels with Industry norms.
4) Specify new requirements.
5) Identify alternative forecasting options.
6) Characterize economic impact of improved forecasts,
7) Select best approaches.
8) Develop implementation schedule.
9) Identify potential problems and bottle neck areas.
10) Implement new system & monitor performance.
Primary object of using Artificial Intelligence is to better integrate Managerial and Quantitative
estimates thereby reducing forecasting errors.
- - -
Comments