Week 1: Overview

Lecture:

Definition of Engineering

Engineering design process

Interdisciplinary Collaboration

Exercise:

Setup of Raspberry Pi

Week 2: Need Assessment

Lecture:

Recognize problems that need an ECE solution

Types of needs: Personal, Financial, needs of humanity, etc.

Work Breakdown, Gantt chart

Exercise:

Problem that needs an ECE solution: Startups tend to fail

Objective: Give a unique score to each startup relating to its chance to fail

Background: Small project that can be accomplished by senior D group with a potential to earn money

Methodology: Begin with a unique data mining system to get important attributes, then develop a ML classifier system

Results: Investors can judge the whether a startup is susceptible to failure

Costs: Very low since it is required as a part of a senior design project

Week 3: Structuring the Search for the Problem

Lecture:

How to formulate a problem statement

KT situation analysis

Techniques to define problems

Exercise:

KT analysis of the problem:

What:

• Is known: public attributes of startups

• Is not known: private attributes and future outcomes

• Are constraints: Time, skill level of creators, low funding

• Are not constraints: Platforms and software used

• Are goals: providing a useful score to weigh the success of a startup

• Are not goals: predicting every future detail perfectly

When:

• When must solution be implemented: significant progress by graduation

• When did changes occur: monthly? quarterly? depends on how often data changes

Who:

• Who can provide more information? Startup experts and websites about them

• Who is the customer? Startup funders

• Who is not the customer? Regular people

• Who is affected by the problem? Investors in new risky companies

• Who is not affected by the problem? Investors in mature, established companies

Where:

• Where is the customer located? Bay Area and other startup hubs

• Where is the customer not located?

Why:

• Why is the problem important? Startups are extremely risky and being able to quantify this risk would make it extremely useful

• Why is the problem not important? Perhaps it is impossible to truly predict these things

• Why does solution work? It can give investors a simple score to make it easier to quantify

• Why does the solution not work? Because it is simplifying a massively complex problem

Week 4: Structuring the Search for a Solution

Lecture:

Subdivision of the problem into design goals

Ergonomic Constraints

Design goals vs. Design specs

Exercise:

Solution development:

• Task 1: Get the data. Develop a unique system to mine important attributes of present and past startups over the years

• Task 2: Design the system. Create a machine learning system that can assign a score from zero to 100 based off the data found in task 1

• Task 3: Analyze results and improve. Train the system until it is able to accurately classify test data.

Week 5: Acquiring, Applying, and Protecting Technical Knowledge

Lecture:

Where to obtain technical information

Patent, trademark, and copyright law basics

Exercise:

My research has shown that many people have done startup prediction for fun as a project on Kaggle but there are no companies, patents or trademarks on that subject yet. It will be looked into in more detail

Many startups have been focused on data analytics of other startups but none have ventured into prediction and machine learning

Week 6: Abstraction and Modeling:

Lecture:

Abstraction

Model symplicity and Occam's razor

Software used in modeling and simulation

Exercise:

Models cannot be created yet due to the need for good data. Once this is available, a model will be made sometime next year. It will be a system mode, not a process model, and a stochastic model, not a deterministic one.

Week 7: Synthesis

Lecture:

Synthesis: how completely new ideas are created

Accidental vs procedural discoveries

Morphological charts

Exercise:

Create morphological chart, in appendix A.

Week 8: Ethics and Product Liability

Lecture:

Safety standards

Liability and legal issues

Data protection

Black Swan and Grey Rhino

Exercise:

Ethical concerns with the startup project

• Is it ethical to mine and store startup data?

• Could it hurt some startups?

• What data is public and what isn't?

Week 9: Failure Analysis

Lecture:

3 Levels:

Level 1: Physical Flaws

Level 2: Process errors

Level 3: Errors in perspective/attitude

Five Whys

Root Cause Analysis

Exercise:

Solution to ethical concerns with startup project

Only use public and legal data

Do not have any bias against certain startups

Week 10:

Lecture:

KT analysis method

Rank-ordering design goals

Eventuality structure

Exercise: KT analysis in Appendix 2

Week 11:

Lecture:

Design for X (DFX)

Purple Cow

Hack vs Kludge

Exercise:

Reviewed the sites listed

Appendix 1: