Model Training for Tiny Devices

Skill id

ai.model-training

Area

Systems & Applications

Group

Edge AI

Roadmap importance

No roadmap colour (added by TESA)

Origin

Topic in the roadmap README

Status in the lesson library

Assessed by a lesson

See it on the skill roadmap

Lessons that develop this skill

Lesson	Course	Develops to
Dataset engineering: class balance, windows and the train/val/test split	Edge AI Developer: From Sensor to On-Device Model	L1 Aware
Training in Docker: one artifact, four targets	Edge AI Developer: From Sensor to On-Device Model	L2 Guided
Inside training: Keras, Conv1D, gradient descent, int8 and the confusion matrix	Edge AI Developer: From Sensor to On-Device Model	L3 Independent
Hands-on: complete the training script and run it in Docker	Edge AI Developer: From Sensor to On-Device Model	L3 Independent

Lessons that assess this skill

Lesson	Level	Evidence
Hands-on: complete the training script and run it in Docker	L2 Guided	practice/s12_train.py

Roles that use this skill

Role	Minimum level	In this role
Edge AI Engineer	L3 Independent	Required (R)

Proficiency levels

• L1 Aware · Bloom: remember/understand
• L2 Guided · Bloom: apply (scaffolded)
• L3 Independent · Bloom: apply/analyse
• L4 Professional · Bloom: analyse/evaluate
• L5 Design & Lead · Bloom: evaluate/create

For Open Badges 3.0 and CASE alignment

UUID: e729c756-a37f-5710-a2fb-b4e5a8a60879

{
  "type": [
    "Alignment"
  ],
  "targetName": "Model Training for Tiny Devices",
  "targetUrl": "https://tesaiot.github.io/tesa-qualification-program/skills/ai.model-training/",
  "targetCode": "ai.model-training",
  "targetFramework": "TESA Embedded Systems Skill Map 0.1.0",
  "targetType": "CFItem"
}

Skill map data is licensed CC BY-SA 4.0, adapted from the Embedded Systems Engineering Roadmap by Meysam Parvizi