[ad_1]
One of the problems business leaders face when communicating with their technical counterparts is describing the AI problem. To simplify some communication, here are some AI problem types.
Try to map AI capabilities to these common problem types. Note that problem types often overlap, but that’s okay. The key is to identify the problem types that most closely match the task at hand when communicating with your AI and data science experts.
Common types of AI problems
1. Classification
A classification The problem is assigning one or more categories to a document, product, person, or image—essentially anything. Examples include:
2. Regression
A The regression problem It is about assessment numeric values for a given input. For example, we try to calculate the number of months until a car needs service given the conditions of the existing machine, or predict how a particular dose of medicine affects blood pressure.
3. Recommendation
A RRecommendation problem is to deliver personalized content or products to a group of people. Examples include:
- Product recommendation
- Recommendations on who to follow
- Recommendations for jobs you should apply for
- Recommendations for articles to read
4. Relevance of the search
A search relevance The problem concerns Improving the ranking of search results displayed to users. Often times, improving search relevancy starts with analyzing search logs, using hard data to diagnose problems. Improving search may or may not require heavy use of machine learning.
or to get information The problem concerns extracting specific information from large volumes of textual data. One of the goals of information mining is to fill patterns using data extracted from raw text. Examples include:
- Extracting patient symptoms from large volumes of clinical records
- Obtaining relevant information from large volumes of legal case materials
- Prefilling candidate application form / database by obtaining relevant information from resumes
6. Summary of the text
Summary of the text is to create an accurate summary of a longer document or set of documents.
7. Clustering
clustering There are people, content, documents, topics, etc. A grouping based on some logical structure, for example, grouping customers according to their purchasing behavior.
More generally, clustering divides data points into several fixed (or dynamic) groups such that data points in one group are more similar to each other than data points in other groups.
9. Virtual AI assistant
A virtual AI assistant Used for short conversations with people to complete simple tasks. Examples include:
- Answering common user questions without human intervention
- Using text messages to check bank balances or make refunds
Alexa and Siri are examples of virtual AI assistants.
10. Mood analysis
sentiment analysis deals with emotion detection in textual data such as user reviews, social media comments, and surveys. For example, automatically detecting customer sentiment in social media channels after a new product launch. Sentiment analysis can even be applied to images to understand emotions from facial expressions.
11. Object detection
Object detection problem Refers to the detection of specific objects in digital images and videos, such as people, buildings, or vehicles.
12. Document segmentation problem
document segmentation There is an attempt to divide the documents into important parts. For example, segmenting unstructured clinical texts to extract their past medical history and family history.
Finding keywords is to identify the terms that best describe the subject of the document – for example, extracting keywords from large volumes of legal documents to understand the topics of discussion.
Although many keyword extraction tools are available (including open source tools), you need to ensure that they work on your data. Often, keyword research tools are best customized or developed.
14. Speech recognition
Speech recognition, also known as speech-to-text (STT) or automatic speech recognition (ASR), is about having a computer program understand and convert spoken language into a written format (or text).
Speech recognition is often used for downstream tasks. For example, speech recognition is used behind the scenes to display relevant search results when you use Google Voice Search. Specifically, your speech is translated into a human-readable format and this generated text is used to display relevant search results.
Many vendors offer speech recognition solutions, so developing speech recognition systems from scratch is rarely necessary. Of course, these systems would benefit from customizing the target data.
15. Machine translation
Machine translation is the automatic software translation of text from one language to another. For example, translating English sentences into German with reasonable accuracy. Machine translation programs rarely need to be developed from scratch, but can benefit from customization.
Machine translation is used for many purposes, including:
- Localization of site text for a specific country
- Customer support chats across countries
- Understanding documents written in another language
Summary of AI problem types
In this short guide, we discuss 15 common types of AI problems that often overlap. For example, you can use a classification approach for sentiment analysis. However, the key is to identify the type of problem that best fits the task at hand. It doesn’t have to be 100% accurate – it’s just semantics. You can continuously refine these definitions with the help of your AI experts.
Keep learning and succeed with AI
- Join my AI Integrated Newsletter, which demystifies AI and teaches you how to successfully deploy AI to drive profitability and growth in your business.
- Read the business case for AI Learn the applications, strategies, and best practices to be successful with AI (select companies using the book: government agencies, automakers like Mercedes Benz, beverage manufacturers, and e-commerce companies like Flipkart).
- Work directly with me Improve your organization’s understanding of AI, accelerate AI strategy development, and get meaningful results from every AI initiative.
Other articles you may be interested in:
connected
[ad_2]
Source link