avatar
Articles
102
Tags
32
Categories
9
Home
About
Showcase
Resume
Danny's Blog
Home
About
Showcase
Resume

Danny's Blog

AWS Certified AI Practitioner(31) - AWS AI Managed Services
Created2025-08-26|CERTIFICATIONAWS_AI_PRACTITIONER
AWS AI Managed ServicesWhy AWS AI Managed Services?AWS AI Managed Services provide pre-trained ML models designed for specific use cases, without requiring you to build or train models from scratch. Key Benefits: Responsiveness and Availability: Always accessible, deployed across multiple Availability Zones and AWS Regions. Redundancy and Reliability: Services remain available even if one AZ experiences downtime. Performance: Use of specialized CPUs and GPUs optimized for ML workloads โ†’ cost ...
AWS Certified AI Practitioner(30) - Hyperparameter Tuning
Created2025-08-26|CERTIFICATIONAWS_AI_PRACTITIONER
Hyperparameter Tuning1. What is a Hyperparameter? Definition: Settings that define how the model is structured and how the learning algorithm works. Set before training begins (they are not learned from the data). Examples: Learning rate Batch size Number of epochs Regularization ๐Ÿ‘‰ Exam Tip: Hyperparameters are not learned during training. They are chosen before training and tuned for best performance. 2. Why Hyperparameter Tuning Matters Goal: Find the best combination of hyperparameter...
(ํ•œ๊ตญ์–ด) AWS Certified AI Practitioner (30) - ํ•˜์ดํผํŒŒ๋ผ๋ฏธํ„ฐ ํŠœ๋‹
Created2025-08-26|CERTIFICATIONAWS_AI_PRACTITIONER_KR
ํ•˜์ดํผํŒŒ๋ผ๋ฏธํ„ฐ ํŠœ๋‹ (Hyperparameter Tuning)1. ํ•˜์ดํผํŒŒ๋ผ๋ฏธํ„ฐ๋ž€? ์ •์˜: ๋ชจ๋ธ ๊ตฌ์กฐ์™€ ํ•™์Šต ๋ฐฉ์‹์„ ๊ฒฐ์ •ํ•˜๋Š” ์„ค์ •๊ฐ’ ํŠน์ง•: ํ•™์Šต์ด ์‹œ์ž‘๋˜๊ธฐ ์ „์— ์ •ํ•ด์ง ๋ฐ์ดํ„ฐ ์ž์ฒด๊ฐ€ ์•„๋‹ˆ๋ผ, ํ•™์Šต ์•Œ๊ณ ๋ฆฌ์ฆ˜์˜ ๋™์ž‘ ๋ฐฉ์‹์— ์˜ํ–ฅ์„ ์คŒ ๋Œ€ํ‘œ ์˜ˆ์‹œ: ํ•™์Šต๋ฅ (Learning rate) ๋ฐฐ์น˜ ํฌ๊ธฐ(Batch size) ์—ํฌํฌ ์ˆ˜(Number of epochs) ์ •๊ทœํ™”(Regularization) ๐Ÿ‘‰ ์‹œํ—˜ ํฌ์ธํŠธ:ํ•˜์ดํผํŒŒ๋ผ๋ฏธํ„ฐ๋Š” ๋ชจ๋ธ ํ•™์Šต ๊ณผ์ •์—์„œ ์ž๋™์œผ๋กœ ํ•™์Šต๋˜๋Š” ๊ฐ’์ด ์•„๋‹ˆ๋ผ, ์‚ฌ์ „์— ์„ค์ •ํ•˜๋Š” ๊ฐ’์ด๋‹ค. 2. ํ•˜์ดํผํŒŒ๋ผ๋ฏธํ„ฐ ํŠœ๋‹(Hyperparameter Tuning) ๋ชฉ์ : ์ตœ์ ์˜ ํ•˜์ดํผํŒŒ๋ผ๋ฏธํ„ฐ ๊ฐ’์„ ์ฐพ์•„ ๋ชจ๋ธ ์„ฑ๋Šฅ์„ ๊ทน๋Œ€ํ™”\ ํšจ๊ณผ: ์ •ํ™•๋„ ํ–ฅ์ƒ ๊ณผ์ ํ•ฉ(Overfitting) ๊ฐ์†Œ ์ผ๋ฐ˜ํ™” ์„ฑ๋Šฅ ๊ฐ•ํ™” ๋ฐฉ๋ฒ•: Grid Search: ๊ฐ€๋Šฅํ•œ ๋ชจ๋“  ์กฐํ•ฉ ํƒ์ƒ‰ Random Search: ์ž„์˜์˜ ์กฐํ•ฉ์„ ํƒ์ƒ‰ ์ž๋™ํ™” ์„œ๋น„์Šค: Amazon SageMaker Automatic Model ...
(ํ•œ๊ตญ์–ด) AWS Certified AI Practitioner (29) - ๋จธ์‹ ๋Ÿฌ๋‹ ํ”„๋กœ์ ํŠธ ๋‹จ๊ณ„
Created2025-08-25|CERTIFICATIONAWS_AI_PRACTITIONER_KR
๋จธ์‹ ๋Ÿฌ๋‹ ํ”„๋กœ์ ํŠธ ๋‹จ๊ณ„ (Phases of Machine Learning Project)1. ๋น„์ฆˆ๋‹ˆ์Šค ๋ชฉํ‘œ ์ •์˜ ๋ชฉํ‘œ: ์–ด๋–ค ๋ฌธ์ œ๋ฅผ ํ•ด๊ฒฐํ• ์ง€ ๋ช…ํ™•ํžˆ ์ •์˜ ์ดํ•ด๊ด€๊ณ„์ž(Stakeholders): ํ”„๋กœ์ ํŠธ์˜ ๊ฐ€์น˜, ์˜ˆ์‚ฐ, ์„ฑ๊ณต ๊ธฐ์ค€์„ ์„ค์ • KPI(ํ•ต์‹ฌ ์„ฑ๊ณผ ์ง€ํ‘œ): ๋ฐ˜๋“œ์‹œ ์ •์˜ํ•ด์•ผ ํ•จ โ†’ ๋ชจ๋ธ์ด ์‹ค์ œ๋กœ ๋น„์ฆˆ๋‹ˆ์Šค ๋ชฉํ‘œ์— ๊ธฐ์—ฌํ•˜๋Š”์ง€ ํŒ๋‹จํ•˜๋Š” ๊ธฐ์ค€ ๐Ÿ‘‰ ์‹œํ—˜ ํฌ์ธํŠธ:๋จธ์‹ ๋Ÿฌ๋‹ ํ”„๋กœ์ ํŠธ์˜ ์ฒซ ๋‹จ๊ณ„๋Š” ํ•ญ์ƒ ๋น„์ฆˆ๋‹ˆ์Šค ๋ฌธ์ œ๋ฅผ ์ •์˜ํ•˜๋Š” ๊ฒƒ. KPI ์„ค์ •์€ AWS ์‹œํ—˜์—์„œ ์ž์ฃผ ๊ฐ•์กฐ๋จ. 2. ๋ฌธ์ œ ์ •์˜์™€ ML ๋ฌธ์ œ๋กœ ์ „ํ™˜ (ML Problem Framing) ๋น„์ฆˆ๋‹ˆ์Šค ๋ฌธ์ œ โ†’ ML ๋ฌธ์ œ๋กœ ๋ณ€ํ™˜ ๋จธ์‹ ๋Ÿฌ๋‹์ด ์ •๋ง ํ•„์š”ํ•œ์ง€, ๋‹ค๋ฅธ ํ•ด๊ฒฐ์ฑ…(์˜ˆ: ๋‹จ์ˆœ ๊ทœ์น™ ๊ธฐ๋ฐ˜)์ด ๋” ๋‚˜์€์ง€ ํŒ๋‹จ ๋ฐ์ดํ„ฐ ๊ณผํ•™์ž, ๋ฐ์ดํ„ฐ ์—”์ง€๋‹ˆ์–ด, ML ์•„ํ‚คํ…ํŠธ, ๋„๋ฉ”์ธ ์ „๋ฌธ๊ฐ€๊ฐ€ ํ•จ๊ป˜ ํ˜‘์—… 3. ๋ฐ์ดํ„ฐ ์ฒ˜๋ฆฌ (Data Processing) ๋ฐ์ดํ„ฐ ์ˆ˜์ง‘ ๋ฐ ํ†ตํ•ฉ: ์ค‘์•™์—์„œ ์ ‘๊ทผ ๊ฐ€๋Šฅํ•˜๋„๋ก ์ •๋ฆฌ ์ „์ฒ˜๋ฆฌ ๋ฐ ์‹œ๊ฐํ™”: ๋ฐ์ดํ„ฐ ํ’ˆ์งˆ ํ™•์ธ, ์ด์ƒ...
AWS Certified AI Practitioner(29) - Phases of a Machine Learning Project
Created2025-08-25|CERTIFICATIONAWS_AI_PRACTITIONER
Phases of a Machine Learning Project1. Define Business Goals Every ML project starts with defining the business objective. Stakeholders must agree on: The value the project will provide The budget The success criteria KPI (Key Performance Indicators) are critical to measure whetherthe ML model actually achieves business goals. ๐Ÿ‘‰ Exam Tip: AWS often asks about the importance of KPIs in framing an ML project. The first step is always business problem definition, not jumping into training a ...
(ํ•œ๊ตญ์–ด) AWS Certified AI Practitioner (28) - ๋จธ์‹ ๋Ÿฌ๋‹ ์ถ”๋ก 
Created2025-08-25|CERTIFICATIONAWS_AI_PRACTITIONER_KR
๋จธ์‹ ๋Ÿฌ๋‹ โ€“ ์ถ”๋ก (Inferencing)1. ์ถ”๋ก ์ด๋ž€? ์ถ”๋ก (Inferencing): ์ด๋ฏธ ํ•™์Šต๋œ ๋ชจ๋ธ์ด ์ƒˆ๋กœ์šด ๋ฐ์ดํ„ฐ์— ๋Œ€ํ•ด์˜ˆ์ธก์„ ๋‚ด๋ฆฌ๋Š” ๊ณผ์ •\ **ํ•™์Šต(Training)**์€ ๋ชจ๋ธ์ด ํŒจํ„ด์„ ๋ฐฐ์šฐ๋Š” ๊ณผ์ •์ด๊ณ ,**์ถ”๋ก (Inferencing)**์€ ํ•™์Šต๋œ ์ง€์‹์„ ํ™œ์šฉํ•˜๋Š” ๋‹จ๊ณ„ 2. ์ถ”๋ก ์˜ ๋‘ ๊ฐ€์ง€ ๋ฐฉ์‹(1) ์‹ค์‹œ๊ฐ„ ์ถ”๋ก  (Real-Time Inference) ๋ฐ์ดํ„ฐ๊ฐ€ ๋“ค์–ด์˜ค๋Š” ์ฆ‰์‹œ ์˜ˆ์ธก์„ ๋‚ด๋ ค์•ผ ํ•˜๋Š” ๊ฒฝ์šฐ ํŠน์ง•: ๋น ๋ฅธ ์†๋„๊ฐ€ ์ค‘์š” (์ •ํ™•๋„๋ณด๋‹ค๋Š” ์†๋„ ์šฐ์„ ) ๊ฒฐ๊ณผ๋ฅผ ์ฆ‰๊ฐ์ ์œผ๋กœ ์ œ๊ณตํ•ด์•ผ ํ•จ ์˜ˆ์‹œ: ์ฑ—๋ด‡, ์Œ์„ฑ ๋น„์„œ(Alexa, Siri), ์˜จ๋ผ์ธ ์ถ”์ฒœ ์‹œ์Šคํ…œ ๐Ÿ‘‰ AWS ์ž๊ฒฉ์ฆ์—์„œ ์ž์ฃผ ๋‚˜์˜ค๋Š” ํฌ์ธํŠธ:์‹ค์‹œ๊ฐ„ ์ถ”๋ก ์€ ์ง€์—ฐ(latency) ์ตœ์†Œํ™”๊ฐ€ ํ•ต์‹ฌ. ๋ชจ๋ธ ์ •ํ™•๋„๊ฐ€ ์กฐ๊ธˆ๋‚ฎ๋”๋ผ๋„ ์ฆ‰๊ฐ์ ์ธ ์‘๋‹ต์ด ํ•„์š”ํ•œ ๊ฒฝ์šฐ ์‚ฌ์šฉ๋จ. (2) ๋ฐฐ์น˜ ์ถ”๋ก  (Batch Inference) ๋Œ€๋Ÿ‰์˜ ๋ฐ์ดํ„ฐ๋ฅผ ๋ชจ์•„์„œ ํ•œ ๋ฒˆ์— ์ฒ˜๋ฆฌํ•˜๋Š” ๋ฐฉ์‹ ํŠน์ง•: ์†๋„๋ณด๋‹ค๋Š” ์ •ํ™•์„ฑ์ด ์ค‘์š” ๋ถ„์„์šฉ์œผ๋กœ ์ฃผ๋กœ ์‚ฌ์šฉ ๊ฒฐ๊ณผ๋ฅผ ๋ฐ›...
AWS Certified AI Practitioner(28) - Machine Learning Inferencing
Created2025-08-25|CERTIFICATIONAWS_AI_PRACTITIONER
Machine Learning โ€“ Inferencing1. What is Inferencing? Inferencing is when a trained model makes predictions on new unseen data. Training = teaching the model. Inferencing = applying what the model has learned to makepredictions. 2. Two Types of Inferencing(1) Real-Time Inference Predictions are made instantly as new data arrives. Key Points: Speed is more important than perfect accuracy. Users expect immediate responses. Examples: Chatbots (customer service bots, Alexa, Siri) Fr...
AWS Certified AI Practitioner(27) - Model Evaluation - Classification & Regression
Created2025-08-23|CERTIFICATIONAWS_AI_PRACTITIONER
๐Ÿ“Š Model Evaluation โ€“ Classification & RegressionWhen building ML models, itโ€™s not enough to just train themโ€”you alsoneed to evaluate how good they are. Different problems (classificationvs regression) use different metrics. Letโ€™s break it down. ๐Ÿ”น Binary Classification Example โ€“ Confusion MatrixA confusion matrix compares actual labels (truth) with the modelโ€™spredictions. True Positive (TP): predicted positive, actually positive\ False Positive (FP): predicted positive, actually negati...
(ํ•œ๊ตญ์–ด) AWS Certified AI Practitioner (27) - ์ด์ง„ ๋ถ„๋ฅ˜์™€ ํ˜ผ๋™ ํ–‰๋ ฌ (Confusion Matrix)
Created2025-08-23|CERTIFICATIONAWS_AI_PRACTITIONER_KR
๐Ÿ“Š Model Evaluation in Machine Learning๋จธ์‹ ๋Ÿฌ๋‹ ๋ชจ๋ธ์„ ๋งŒ๋“ค์—ˆ์„ ๋•Œ, ์„ฑ๋Šฅ์ด ์ž˜ ๋‚˜์˜ค๋Š”์ง€๋ฅผ ํ™•์ธํ•˜๋Š” ๊ณผ์ •์ด ํ•„์š”ํ•ฉ๋‹ˆ๋‹ค.์ด๋•Œ ๋ถ„๋ฅ˜(Classification) ๋ชจ๋ธ๊ณผ ํšŒ๊ท€(Regression) ๋ชจ๋ธ์˜ ํ‰๊ฐ€ ๋ฐฉ์‹์ด ๋‹ค๋ฅด๋ฏ€๋กœ ๊ตฌ๋ถ„ํ•ด์„œ ์•Œ์•„๋‘์–ด์•ผ ํ•ฉ๋‹ˆ๋‹ค. ๐Ÿ”น ์ด์ง„ ๋ถ„๋ฅ˜ (Binary Classification)์™€ ํ˜ผ๋™ ํ–‰๋ ฌ (Confusion Matrix)Confusion Matrix๋ž€? ์‹ค์ œ ์ •๋‹ต(๋ผ๋ฒจ)๊ณผ ๋ชจ๋ธ ์˜ˆ์ธก๊ฐ’์„ ๋น„๊ตํ•ด์„œ ์„ฑ๋Šฅ์„ ํ‰๊ฐ€ํ•˜๋Š” ๋„๊ตฌ ๋„ค ๊ฐ€์ง€ ๊ฐ’์œผ๋กœ ๋‚˜๋‰ฉ๋‹ˆ๋‹ค: ๊ตฌ๋ถ„ ์˜ˆ์ธก Positive ์˜ˆ์ธก Negative ์‹ค์ œ Positive True Positive (TP) False Negative (FN) ์‹ค์ œ Negative False Positive (FP) True Negative (TN) ๐Ÿ‘‰ ๋ชฉํ‘œ: TP์™€ TN์„ ์ตœ๋Œ€ํ™”ํ•˜๊ณ , FP์™€ FN์„ ์ตœ์†Œํ™”ํ•˜๋Š” ๊ฒƒ. ์ฃผ์š” ํ‰๊ฐ€ ์ง€ํ‘œ (Classification...
(ํ•œ๊ตญ์–ด) AWS Certified AI Practitioner (26) - ๋ชจ๋ธ ์ ํ•ฉ๋„์™€ ํŽธํ–ฅ & ๋ถ„์‚ฐ
Created2025-08-23|CERTIFICATIONAWS_AI_PRACTITIONER_KR
๐Ÿค– ๋ชจ๋ธ ์ ํ•ฉ๋„(Model Fit)์™€ ํŽธํ–ฅ(Bias) ยท ๋ถ„์‚ฐ(Variance)1. ๋ชจ๋ธ ์ ํ•ฉ๋„(Model Fit)๋จธ์‹ ๋Ÿฌ๋‹ ๋ชจ๋ธ์ด ์ œ๋Œ€๋กœ ๋™์ž‘ํ•˜์ง€ ์•Š์„ ๋•Œ๋Š” ๋ชจ๋ธ์˜ ์ ํ•ฉ๋„(Fit) ๋ฅผ ์‚ดํŽด๋ด์•ผ ํ•ฉ๋‹ˆ๋‹ค.๋ชจ๋ธ์ด ๋ฐ์ดํ„ฐ๋ฅผ ์–ผ๋งˆ๋‚˜ ์ž˜ ์„ค๋ช…ํ•˜๋Š”์ง€๊ฐ€ ํ•ต์‹ฌ์ž…๋‹ˆ๋‹ค. ๊ณผ์ ํ•ฉ(Overfitting) ํ›ˆ๋ จ ๋ฐ์ดํ„ฐ์—์„œ๋Š” ์„ฑ๋Šฅ์ด ๋งค์šฐ ์ข‹์Œ ์ƒˆ๋กœ์šด ๋ฐ์ดํ„ฐ(๊ฒ€์ฆ/ํ…Œ์ŠคํŠธ ๋ฐ์ดํ„ฐ)์—์„œ๋Š” ์„ฑ๋Šฅ์ด ๋‚˜์จ ์›์ธ: ๋ชจ๋ธ์ด ๋ฐ์ดํ„ฐ์˜ ๋…ธ์ด์ฆˆ๊นŒ์ง€ ํ•™์Šตํ•ด์„œ ์ผ๋ฐ˜ํ™”๊ฐ€ ์•ˆ ๋จ ๐Ÿ“Œ ์˜ˆ์‹œ: ํ›ˆ๋ จ ๋ฐ์ดํ„ฐ ์  ํ•˜๋‚˜ํ•˜๋‚˜์— ๋งž๊ฒŒ ์„ ์„ ๊ตฌ๋ถ€๋ ค ๋งŒ๋“  ๋ณต์žกํ•œ ๊ณก์„  ๊ณผ์†Œ์ ํ•ฉ(Underfitting) ํ›ˆ๋ จ ๋ฐ์ดํ„ฐ์—์„œ๋„ ์„ฑ๋Šฅ์ด ๋‚˜์จ ์›์ธ: ๋ชจ๋ธ์ด ๋„ˆ๋ฌด ๋‹จ์ˆœํ•˜๊ฑฐ๋‚˜, ํŠน์ง•(Feature)์ด ๋ถ€์กฑํ•จ ๐Ÿ“Œ ์˜ˆ์‹œ: ๋ณต์žกํ•œ ๊ณก์„  ๋ฐ์ดํ„ฐ์— ๋‹จ์ˆœ ์ง์„ ์„ ์–ต์ง€๋กœ ์ ์šฉ ๊ท ํ˜•(Balanced) ๊ณผ์ ํ•ฉ๋„, ๊ณผ์†Œ์ ํ•ฉ๋„ ์•„๋‹Œ ์ƒํƒœ ์–ด๋А ์ •๋„ ์˜ค์ฐจ๋Š” ์žˆ์ง€๋งŒ, ๋ฐ์ดํ„ฐ์˜ ์ „์ฒด์ ์ธ ํŒจํ„ด์„ ์ž˜ ๋”ฐ๋ฆ„ ๐Ÿ“Œ ๊ฐ€์žฅ ์ด์ƒ์ ...
1…345…11
avatar
Danny Ki
A data engineer's journey in coding, analytics, and building real-world systems.
Articles
102
Tags
32
Categories
9
Follow Me
Announcement
This is my Blog
Recent Posts
Coding_Quiz_SQL_12025-09-17
Coding_Quiz_Python_12025-09-17
Databricks CV Anomaly Detection2025-09-15
(ํ•œ๊ตญ์–ด) AWS ML Associate (6) - Amazon S3 ํ•ต์‹ฌ ์ •๋ฆฌ2025-09-14
(ํ•œ๊ตญ์–ด) AWS ML Associate (5) - Amazon S3 ํ•ต์‹ฌ ์ •๋ฆฌ2025-09-14
Categories
  • CERTIFICATION88
    • AWS_AI_PRACTITIONER41
    • AWS_AI_PRACTITIONER_KR41
    • AWS_ML_ASSOCIATE_KR6
  • Dev7
    • Algorithm5
    • CODING_QUIZ_PYTHON1
    • CODING_QUIZ_SQL1
Tags
Machine Learning AWS Python Apache_Flink Databricks OpenAI Quick Sort DP Insertion Sort Bubble Sort SQL Kaggle Grafana MLflow XGBoost NLP Elasticsearch Apache Spark Dynamic Programming PYTHON LEETCODE Selection Sort AWS_ML_ASSOCIATE AWS_AI_PRACTITIONER Deep Learning Regression PostgreSQL Kafka Computer Vision KOREAN Sorting Swift
Archives
  • September 2025 29
  • August 2025 66
  • July 2025 7
Website Info
Article Count :
102
Unique Visitors :
Page Views :
Last Update :
© 2025 By Danny KiFramework Hexo 7.3.0|Theme Butterfly 5.4.3