Starting from watching the movie A.I., I fancied about how computational machine will evolve into something similar to human brain with logical thinking, abundant knowledge and good attitude in learning new things around.
You might also have witnessed the horror originated from another movie like Terminator. All in a sudden, a killer machine might jump out from nowhere to search for its next target for termination.
Before we can judge what A.I. can do good or evil to human mankind, we'd better learn more about what exactly is A.I..
In a business sense, Artificial Intelligence is a broad concept about manipulating computing resources in order to deliver services and business intelligence to both customers and business owner. The target customers here refers to human on Earth, Mars or within the entire galaxy. When there's human, there's business. To do a successful business with human, it's crucial to deliver the services in a customised way.
Here we don't tell anything about the good or evil side of A.I.. As we all know, the technology cannot do bad things. It's just human who uses it to do something good or evil. Even humans think they are doing good things, it may end up the consequence is not what they desire, in other words, bad thing comes.
Within the concept of A.I., we have Machine Learning. M.L. is the business process to consume existing data collected from the market. In other words, a neural network can access and analyse the business data whether it's structural or non-structual. The outcome can be fruitful as we all have witnessed the success of Google and Facebook about making use of A.I. to deliver better functionalities and features.
Within Machine Learning, we have Deep Learning or Deep Machine Learning or deep structured learning or hierarchical learning. They refer to the same thing, a set of Machine Learning Algorithms aiming to extract a higher level of meanings from the sea of data collected from the market. Simple things like who the customers are, where they're residing, what interests do they have can drive amazingly new direction in providing better services to fit the individual's needs. Here we stressed the individual, it's everyone. Fitting everyone's need seems to be impossible for the firm or the corporate in the last century. To tailor make the needs for everyone, we need to collect huge amount of data related to customer's behaviour and interest and then carry out a series of analysis to extract beneficial instructions to make our services suitable for the loyal customers.
When we talk about Deep Learning, we will also need to mention about artificial neural network which is where all those Machine Learning algorithms running within. There are many types of neural networks invented so far.
Here comes a list of typical neural networks:
- Group Method of Data Handling (GMDH)
- Convolutional neural networks
- Neural history compressor
- Recursive neural networks
- Long short-term memory (LSTM)
- Deep belief networks
- Convolutional deep belief networks
- Large memory storage and retrieval neural networks
- Deep Boltzmann machines
- Stacked (de-noising) auto-encoders
- Deep stacking networks
- Tensor deep stacking networks
- Spike-and-slab RBMs
- Compound hierarchical-deep models
- Deep coding networks
- Deep Q-networks
- LSTM-related differentiable memory structures
- Semantic hashing
- Neural Turing machines
- Memory networks
- Pointer networks
- Encoder–decoder networks
- Multilayer kernel machine
For nowadays Deep Learning technology, here's the list of applications:
- Automatic speech recognition
- Image recognition
- Natural language processing
- Drug discovery and toxicology
- Customer relationship management
- Recommendation systems
- Biomedical Informatics
They are not purely in a business sense but we can borrow these ideas and generate something new. Here's the refined list with a sense of business:
- Customer speech recognition
- Customer facial recognition
- Human language processing for different countries and dialects
- Market discovery and SWOT analysis based on big data
- Customer relationship management
- Service recommendation systems
- Gene discovery relating to customer behaviour and choice
Of course, there are many more to come. Yet we take a break to digest all these topics above and see if we can take one more step towards the A.I. business.
