Last month in the column entitled "Digital Business and Your Digital Intelligence Maturity," my associate posed the question: Are you ready for the world of digital business? In that column, four implementation approaches for digital intelligence were discussed.
This month and next, I will go into more detail in the area described as emerging markets experts. This type of solution is developed based on understanding industry trends, neural networks and predictive analysis. The key is to position your organization at the leading edge of the next wave of solutions. The use of neural networks and predictive analysis is a way to move ahead of competition.
How Neural Networks Work
Neural networks are named after the cells within the human brain that perform intelligent operations. The brain is made up of billions of neuron cells. Each of these cells is like a tiny computer with extremely limited cap-abilities; however, connected together, these cells form the most intelligent nervous system known. Neural networks are formed from hundreds or thousands of simulated neurons connected together in much the same way as the brain's neurons.
Just like people, neural networks learn from experience, not from programming. Neural networks are proficient and capable of pattern recognition, generalization and trend prediction. They are fast, tolerant of imperfect data and do not need formulas or rules. By repeatedly presenting examples to the network, users train neural networks. Each example includes both inputs (information one would use to make a decision) and outputs (the resulting decision, prediction or response).
Your network tries to learn each of your examples in turn, calculating its output based on the inputs you provided. If the network output doesn't match the target output, there are programs that can correct the network by changing its internal connections. This trial-and-error process continues until the network reaches your specified level of accuracy. Once the network is trained and tested, you can give it new input information, and it will produce a prediction. Designing your neural network is largely a matter of identifying which data is input and what you want to predict, assess, classify or recognize.
Neural Network Components
Neural networks are made up of three different layers:
- Input: The input layer is connected to the meta data.
- Processing: The middle processing layers are used to consolidate the results of the input layer. They are made up of many nodes, which simulate neurons by their interconnection to other nodes.
- Output: The output layer produces the results.
Features and Benefits
Neural networks are excellent point solution tools where the need for pattern recognition is necessary, such as discovering customer spending patterns in a database, but they are not good knowledge transfer tools because they have no external representation of what they know. The following are some questions to ask when considering using neural networks:
- Adaptability: Can the technology learn from its environment? In the case of neural networks, learning from data is a primary attribute.
- Flexibility: Can the technology represent multiple types of knowledge? In the case of neural networks, they can only represent pattern recognition knowledge but do not reveal explicit detail about what they know to the user.
- Transferability: Does the knowledge representation associated with the technology make it easy for people to learn from the representation? Because neural networks have no way to make what they know explicit, they can only teach through continued interaction, with people applying their own pattern recognition skills to attempt to glean what the network has inferred, resulting in a very low transferability rating for neural networks.
Next month, I will discuss designing neural networks and current and future trends. If you desire to learn more about neural networks, there are several good books on the market discussing this topic. One of the most recently published books on this subject is Beyond the Internet Alien Intelligence by Dr. James Martin. This provides very good insight into the use of neural networks.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access