Information
Leave a reply
Category Archives: Uncategorized
AI Beneath the covers!
Artificial intelligence as has four basic subject areas. They are as follows:
Statistics or, features
Linear algebra
Calculus
Probability
I would like to focus on statistics or AI eatures.
Information including statistics are derived from data ,data is real potential. It’s not calculated or surprised that will come later. And it most probably reflects some type of transaction
Information which is derived from data, can be standardized by reference data, or calculated of cours,e
Features can also be transformed, standardized and, consist of calculated information and be in many cases, change from its raw form
Then, of course we have to consider the algorithm you may choose to predict or especially recently todays generative content
It is also worth considering that in the early AI days much data concerns came from structured , sources, tables, or dimensions and columns, or as a attributes this data, what is the model and kept in and RDBMS recently we are able to process mature what we would call a structured data. This presents both great opportunity, and a greater chance for not understanding the context of the data.
My main conclusion is that you have to be careful that you consume data or feed it to your AI with a forensically traceable and verifiable source. In other words, he a chain of custody as you would in the real world. You cannot just take a prediction and not understand how it was. Arrived at. Like in the real world, there’s a difference between a guess and a well thought out prediction
Something you may want to know before you go that way , We have yet learned how our judgment works. Apparently, it’s a mixture of learned knowledge, experiences in human intelligence
Data Mind Set interview Information Value Chain
Pending Data Apocalypse
Artificial Intelligence Codex
This is a brief blog to help understand the synergies and similarities between AI & BI We will start our at what drives the process, which is the business goals I will be publishing sevral guides and templates\ to help you in this but is important this is written for your team and others to evaluat
Businemss Intelligence
This path will involve recording and providing the deliverables so let your development team translate your requirements into business and technical deliverables
Equally we will use these deliverables as a springboard in defining, creating and documenting interesting business and technical deliverables into functional deliverables from the business that the AI development team can use
BI Dimensionns
BI dimensions is it area we will cover quite a bit. It is similar here distinctly different from AI use of the term in PI. A dimension is a high-level category or a category of glossary terms or attitude from a file or table, bottom line in BI. It’s a grouping of separate fields or an AI each field is called equally a feature, or a dimension, or a variable. In AI it is around the algorithm selected. We will cover this in detail.
BI Metrics
AI measure a KPI or metric that is piece part of many data elements and several parts of dimensions which is the name of the dimension or category in business terms
AI Features
In AI features is interesting, and a bit different than the grooving of separate fields or data elements or attributes, which is called business dimensions. A feature has several definitions. The primary definition is a field that that in part of an algorithm chosen and impacts the algorithm based on the input in AI. There are many techniques and tools to.A high features is interesting, and a bit different than the in AI there are many tools and techniques for feature selection. They are very similar and based on the tools and techniques used in BIA to define attributes who will discuss in detail later.grooving of dittos, which is called business dimensions. A feature has several definitions. The primary definition is a field that can choose or impacts the model based on the input in AI. There are many techniques and tools to.in AI there are many tools and techniques for feature selection. They are very similar and based on the tools and techniques used in BI to define attributes who will discuss in detail later.
AI predicted Content
Here too, we have not standardized on description. The predicted content is commonly known as the generated content. this is important in both models types, both expert AI, and knowing that there are features, however, LLM specifically creates generated content, or its prediction of your contest based on your questions or prompts
AI & BI Synthesis
I created a an approach and a short video, sharing the best of both worlds and suggesting the best approach to get the best of both worlds
Here is the video
Data Governance NOT Data Deciphering Instead
Video explaining deciphering, which is in principle the logical concept that is used extensively in AI,LLMand ChatGPT In my next Post will clarify  in detail to AI and BI also ChatGPT , LLM and their use/reuse at the business and technical level
Practical artificial intelligence and it’s application to traditional business intelligence
There are many books and photos that talk about the so-called revolution of artificial intelligence or AI I appreciate their enthusiasm however, I don’t think that is really being recognized that much of what they have invented i can be used in reality to greatly, reduce the time and effort to creat the kind of things we’ve had a creat for decades And they are essential to the creation of AI, which is predicated on learning and contact
One of the things that we will address is, there is a proclivity to use much language much of it is not very semantic. Today folks are describing architecture that already have descriptions. We have a mix of marketing terms and creative terms that mean the same thing, but they use words from the past. It’s causing confusion.
Just one example, maybe the term dimensionis one thing it meant to me thing 10 years ago come today it’s being used in a different context, in AI is it’s dimension or parameter or a feature There are many people of many years in a IT language is important it’s semantic
Another is the term similarity . It is used completely different in AI versus traditional, fuzzy matching. true the concept is the same but the Technical use Is different
There is no doubt of the benefit of what’s been created through the use of neural networks and transformers tthat hey can have tremendous positive impact on delivering business intelligence with the aid of artificial intelligence, machine, learning, and Deep Learning subsequently
I have been deeply involved in business intelligence, data quality, data profiling, and MDM and Data Governance for several decades.
I would like to take you on a journey and be able to help you exploit all these capabilities today and yesterday we are experience in the evolution of what we’re doing it is not a revolution. It is an evolution. if anything I hope to help achieve a basic understanding and terminalnology used in information architecture and various techniques that we have that will help , frankly nobody has a corner on the best approach it has all been done before at the logical level I want to be a part of helping us leverage, reuse, and apply what we were doing for decades, to what is now being introduced in the last several years you have to judge among the three goals, cheaper faster, better or we can guarantee is cheaper and faster. It’s up to you to make it better not necessarily the technology.
More to come . . .p
AI LLM New Project Roles including Prompt Engineering and Role AI Scientist and integrated data and information
I’d like to offer some advice on transitioning skills and knowledge skills and knowledge who worked hard to retainto include some of the new AI and NLM developments it’s actually less impactful, and better than you may think
Well, I dressed up in a little bit for now let’s talk about prompt engineers You most likely have currenb SME’s or expert on your current data and requirements
First, do you notice I used the term AI scientist Instead of data scientist , A data scientist Currently is actually a AI model scientist and they will help you apply. We are concerned here with how a lot of folks have opinions heuristic no not necessarily fact-based already we are going to suggest some techniques, and provide some mentoring to explore this important factor in AI is proper training we specialize in providing techniques and mentoring in separate information which is Formulated and an opinion and facts or Data, which, cannot change
There is a series of steps involved in preparing for the use of Data inAI and Chat in the form of LLM models. This is not much different and you may have most of the information already gathered in order to properly design the requirements for your model we would collect the phone.it is important to realize the steps are critical, for you have confidence in your models putput which will be your result of integrating, your Word documents, your Presentations,, spreadsheets, and, of course your actual data.
We wiKeaton, Billy, and modeling of information words versus modeling and requirements for data preparation . There is a difference that is extremely important and in line with what you’ve been doing.I know that Data preparation is not glamorous, but in my 20+ years you’ll get nowhere without proper data preparation you can’t AI it you can’t impute it you need to discuss requirements with people and write them down and then execute it The AI will make the legwork, faster. but in the end you’ll have to review it otherwiseit otherwise you may end up needlessly retracing your steps based on Improper preparation I know that Data preparation is not glamorous, but in my 20+ years you’ll get nowhere without proper data preparation you can AI it you can’t imputed you need to discuss requirements with people and write them down and then execute a I will do it. Faster time is the legwork, but in the end you’ll have to review with Stuck you may end up needlessly retracing your steps based on.I know that Data preparation is not glamorous, but in my 20+ years you’ll get nowhere without proper data preparation you can AI it you can’t imputed you need to discuss requirements with people and write them down and then execute a I will do it. Faster time is the legwork, but in the end you’ll have to review with Stuck you may end up needlessly retracing your steps based on. improper preparation. This can be at Floydd by phone, the proper steps.
1. Word document
2. Presentations
3. Spreadsheets
4.Data reports
5. Data quality report for AI preparation
6.Internet
7.Other Sources (Network,Internet or Local)
We have suggested tools/techniques/open source.and suggestions for each of these. Don’t let that bother you, however, is important with today’s capabilities in AI integrate your words your thoughts, your abstraction, and your actual data together in order for you. They’re trustworthy results from your AI.
We will be providing a separate post on each of these and then finally how they come together our point is that the what you’ve been doing to understand and form requires for tradition BI can be reutilized and extend it for AI
Ira Warren Whiteside's Blog- Information Sherpa 