Research
It is an art
Or is it?
Seems to be the
things I discover
questions I raise
evidence I present
context I find
tests I prove true
Theories which further augment
Curiosity satisfied
Developments pushed further
I’m not here to write a poem but I have a creative background and it’s about perspective sometimes not just 1+1=1
I believe this will reveal its self along the way interestingly and innovate somehow
But let me explain more about the rubix in time!
100+ fields of Ai exist - it’s continually expanding so I’m not upto date in it all - neither can you even be, so none of us know
ok! Now …
Some areas in Ai (which I’ve casually researched around other things since 2020) include:
• KRR
• CV
• Scene reconstruction
• Object detection
• Event detection
• Activity recognition
• Video tracking
• Object recognition
• 3D pose estimation
• Learning
• Indexing
• Motion estimation
• 3D scene modeling
• Image recognition
• NLP
• Automatic Reasoning
• Interdisciplinary Fields
• linguistics
• Automatic reasoning
• Expert Systems (ES)
• Fuzzy logic
• Apps in ctrl theory & AI
• Evolutionary Computation
• Speech Processing
• NLU/NLG
• AR/ VR
• Planning
• RL
• Swarm Intelligence
• Game AI
GenAI
• Content Gen
• Text gen
• Image gen
• Music gen
• Code gen
LLM
• Natural Language Generation (NLG)
• ChatGPT
• GPT
• GPT-4
• Advanced Topics
• Quantisation
• Reward mechanisms
• New LLM architectures
• Domain-specific models
Multimodal Gen Models
• Vision-Language Models (VLMs)
Advanced Retrieval-Augmented Generation (RAG)
• Effective Data Augmentation Techniques
• Large Knowledge Bases
• Refining Retrieval Mechanisms
ML
• Algorithms that Learn
• ANNs
• Predictive Analytics
• Business Analytics applying ML
DL
• Advanced ML Techniques Using ANNs
• Long Short-Term Networks (LSTN)
• Deep Belief Networks (DBN)
• Recurrent Attention Networks (RAN)
• Generative Adversarial Networks (GAN)
• Other
This structured breakdown organises the vast field of AI into specific disciplines & subsections
As mentioned it’s not fully comprehensive
Part of the exploration is Quantum Machine Learning: Exploration of how quantum computing can enhance machine learning algorithms and processes, potentially improving efficiency and performance in data analysis and pattern recognition.
- Parts of my Folio will be linked eventually to better understand design, Ux, theories, test, build, ethics & audit etc
LLMs like GPT-4 and Claude are considered for deep contextual understanding because of their large training data, reinforcement learning from human feedback (RLHF).
They are often paired with multi-agent systems in CrewAI
Beyond this research check out the areas few venture:https://dev.to/gracerosen/off-the-beaten-track-1c4d
Top comments (0)