FIN510 Assessments

Complete assessment guide and resources

1 Assessment Overview

FIN510 uses a three-component assessment structure designed to evaluate both theoretical understanding and practical implementation skills.

1.1 Assessment Summary

Component	Weight	Format	Timing	Feedback
Coursework 1	34%	2 × MCQ Tests (17% each)	Weeks 5 & 12	Following day
Coursework 2	16%	Case Study Analysis	Week 7	+2 weeks
Coursework 3	50%	Python Project Report	January 14th	+3 weeks

1.2 Coursework 1: Multiple Choice Tests (34%)

1.2.1 Test 1 - Week 5 (17%)

Coverage: Python Foundations (Weeks 1-4)

Topics Include: - Python programming fundamentals for finance - Financial data acquisition and APIs - Time series analysis basics - Statistical analysis and risk metrics - Data visualization techniques

Format: 40 minutes, closed book, via Blackboard

1.2.1.1 Sample Questions:

Question 1: Which Python method correctly calculates log returns?

(price_t / price_t-1) - 1
np.log(price_t / price_t-1) ✓
(price_t - price_t-1) / price_t-1
price_t - price_t-1

Explanation: Log returns are calculated as the natural logarithm of the price ratio, which provides better statistical properties for financial analysis.

Question 2: What does a Sharpe ratio of 1.5 indicate?

150% annual return
1.5 units of excess return per unit of risk ✓
1.5% volatility
$1.50 profit per dollar invested

Explanation: The Sharpe ratio measures risk-adjusted returns by dividing excess return by volatility.

1.2.2 Test 2 - Week 12 (17%)

Coverage: AI and Advanced Applications (Weeks 6-11)

Topics Include: - Machine learning model evaluation - Natural language processing in finance - AI ethics and bias considerations - Automated trading systems - Production deployment concepts

1.3 Coursework 2: Case Study Analysis (16%)

1.3.1 Format

Duration: 40 minutes
Type: Open book
Location: Computer lab
Week: 7

1.3.2 Structure

Part A (50%): 20 multiple choice questions
Part B (50%): Written analysis of provided dataset

1.3.3 Sample Case Study

Scenario: You are a data scientist at a fintech startup developing a credit scoring model. You have been provided with results from three different machine learning models trained on lending data.

Your Task: Evaluate the models and recommend the best approach for production deployment.

Provided Data: - Model performance metrics (accuracy, precision, recall, AUC) - Feature importance rankings - Bias audit results - Computational requirements

Analysis Requirements: 1. Compare model performance across different metrics 2. Assess potential bias and fairness issues 3. Consider regulatory compliance requirements 4. Recommend deployment strategy with justification

1.3.3.1 Assessment Criteria:

Technical Understanding (40%): Correct interpretation of ML metrics
Critical Analysis (30%): Evaluation of trade-offs and limitations
Business Application (20%): Practical deployment considerations
Communication (10%): Clear and professional presentation

1.4 Coursework 3: Final Project (50%)

1.4.1 Project Requirements

Report: 2,000 words maximum
Code: Jupyter notebook with implementation
Due Date: January 14th, 2026 by 12:00
Submission: Turnitin + GitHub repository

1.4.2 Project Options

1.4.2.1 Option 1: AI-Powered Trading Strategy

Objective: Develop a complete algorithmic trading system

Requirements: - Implement multiple ML models for price prediction - Create backtesting framework with realistic constraints - Build performance analytics dashboard - Address risk management and regulatory considerations

Deliverables: - Trading strategy implementation - Backtesting results and analysis - Risk assessment framework - Performance comparison with benchmarks

1.4.2.2 Option 2: Financial NLP Application

Objective: Build sentiment analysis system for financial markets

Requirements: - Multi-source sentiment analysis (news, social media, earnings calls) - Correlation analysis with market movements - Real-time monitoring dashboard - Model validation and performance evaluation

Deliverables: - NLP pipeline implementation - Sentiment-price correlation analysis - Interactive monitoring dashboard - Business case and ROI analysis

1.4.2.3 Option 3: Robo-Advisory Platform

Objective: Create automated portfolio management system

Requirements: - Risk profiling questionnaire and algorithm - Modern Portfolio Theory implementation - Rebalancing and tax optimization - Client reporting and communication system

Deliverables: - Complete robo-advisory system - Portfolio optimization algorithms - Client interface and reporting - Regulatory compliance framework

1.4.2.4 Option 4: Credit Risk AI System

Objective: Develop ML-based credit assessment platform

Requirements: - Advanced feature engineering with alternative data - Multiple ML model comparison - Bias detection and fairness algorithms - Model governance and monitoring framework

Deliverables: - Credit scoring model implementation - Bias audit and mitigation strategies - Model monitoring dashboard - Regulatory compliance documentation

1.4.3 Assessment Criteria

1.4.3.1 Technical Implementation (40%)

Code Quality: Clean, well-documented, reproducible code
Algorithm Selection: Appropriate choice of methods and techniques
Data Handling: Proper preprocessing and validation
Performance: Model accuracy and computational efficiency

1.4.3.2 Business Application (25%)

Problem Definition: Clear business case and objectives
Implementation Strategy: Realistic deployment considerations
Risk Assessment: Identification and mitigation of key risks
Value Proposition: Clear articulation of business benefits

1.4.3.3 Critical Analysis (20%)

Literature Review: Integration of relevant academic and industry sources
Methodology Justification: Rationale for chosen approaches
Limitations Discussion: Honest assessment of model constraints
Alternative Approaches: Consideration of other possible solutions

1.4.3.4 Communication (15%)

Report Quality: Professional writing and presentation
Visualization: Effective charts, graphs, and dashboards
Documentation: Clear technical documentation
Referencing: Proper academic citation format

1.5 Assessment Support

1.5.1 Preparation Resources

Sample Projects: Available on course GitHub
Code Templates: Starter notebooks for each project type
Rubric Details: Comprehensive marking criteria
Past Examples: Anonymized high-quality submissions

1.5.2 Getting Help

Office Hours: Weekly consultation sessions
Peer Review: Structured feedback sessions
Technical Support: Lab assistants for coding issues
Writing Support: University writing center resources

1.5.3 Submission Guidelines

File Naming: SurnameFirstNameBNumber_ProjectTitle
Code Repository: GitHub with clear README
Report Format: Word document via Turnitin
Late Penalties: As per university policy

1.6 Academic Integrity

1.6.1 Permitted Collaboration

Discussion: General concepts and approaches
Code Review: Peer feedback on implementation
Resource Sharing: Publicly available datasets and libraries

1.6.2 Prohibited Activities

Code Copying: Submitting someone else’s implementation
Report Plagiarism: Copying text from any source without attribution
Data Fabrication: Creating fake results or datasets

1.6.3 AI Tools Policy

ChatGPT/Copilot: Permitted for learning and debugging
Code Generation: Must be acknowledged and understood
Report Writing: AI assistance must be disclosed
Final Responsibility: You must understand and defend all submitted work

1.7 Grade Boundaries

Classification	Percentage Range	Description
First Class	70-100%	Outstanding work demonstrating innovation and mastery
Upper Second	60-69%	Good quality work with solid understanding
Lower Second	50-59%	Acceptable work meeting basic requirements
Third Class	40-49%	Adequate work with limited demonstration of skills
Fail	0-39%	Insufficient demonstration of learning outcomes

For detailed rubrics and additional support materials, see the Course Handbook.

--- title: "FIN510 Assessments" subtitle: "Complete assessment guide and resources" --- # Assessment Overview FIN510 uses a three-component assessment structure designed to evaluate both theoretical understanding and practical implementation skills. ## Assessment Summary | Component | Weight | Format | Timing | Feedback | |-----------|--------|--------|---------|----------| | **Coursework 1** | **34%** | 2 × MCQ Tests (17% each) | Weeks 5 & 12 | Following day | | **Coursework 2** | **16%** | Case Study Analysis | Week 7 | +2 weeks | | **Coursework 3** | **50%** | Python Project Report | January 14th | +3 weeks | --- ## Coursework 1: Multiple Choice Tests (34%) ### Test 1 - Week 5 (17%) **Coverage**: Python Foundations (Weeks 1-4) **Topics Include:** - Python programming fundamentals for finance - Financial data acquisition and APIs - Time series analysis basics - Statistical analysis and risk metrics - Data visualization techniques **Format**: 40 minutes, closed book, via Blackboard #### Sample Questions: ::: {.sample-question} **Question 1**: Which Python method correctly calculates log returns? A) `(price_t / price_t-1) - 1` B) `np.log(price_t / price_t-1)` ✓ C) `(price_t - price_t-1) / price_t-1` D) `price_t - price_t-1` *Explanation*: Log returns are calculated as the natural logarithm of the price ratio, which provides better statistical properties for financial analysis. ::: ::: {.sample-question} **Question 2**: What does a Sharpe ratio of 1.5 indicate? A) 150% annual return B) 1.5 units of excess return per unit of risk ✓ C) 1.5% volatility D) $1.50 profit per dollar invested *Explanation*: The Sharpe ratio measures risk-adjusted returns by dividing excess return by volatility. ::: ### Test 2 - Week 12 (17%) **Coverage**: AI and Advanced Applications (Weeks 6-11) **Topics Include:** - Machine learning model evaluation - Natural language processing in finance - AI ethics and bias considerations - Automated trading systems - Production deployment concepts --- ## Coursework 2: Case Study Analysis (16%) ### Format - **Duration**: 40 minutes - **Type**: Open book - **Location**: Computer lab - **Week**: 7 ### Structure - **Part A (50%)**: 20 multiple choice questions - **Part B (50%)**: Written analysis of provided dataset ### Sample Case Study ::: {.case-study} **Scenario**: You are a data scientist at a fintech startup developing a credit scoring model. You have been provided with results from three different machine learning models trained on lending data. **Your Task**: Evaluate the models and recommend the best approach for production deployment. **Provided Data**: - Model performance metrics (accuracy, precision, recall, AUC) - Feature importance rankings - Bias audit results - Computational requirements **Analysis Requirements**: 1. Compare model performance across different metrics 2. Assess potential bias and fairness issues 3. Consider regulatory compliance requirements 4. Recommend deployment strategy with justification ::: #### Assessment Criteria: - **Technical Understanding (40%)**: Correct interpretation of ML metrics - **Critical Analysis (30%)**: Evaluation of trade-offs and limitations - **Business Application (20%)**: Practical deployment considerations - **Communication (10%)**: Clear and professional presentation --- ## Coursework 3: Final Project (50%) ### Project Requirements - **Report**: 2,000 words maximum - **Code**: Jupyter notebook with implementation - **Due Date**: January 14th, 2026 by 12:00 - **Submission**: Turnitin + GitHub repository ### Project Options #### Option 1: AI-Powered Trading Strategy **Objective**: Develop a complete algorithmic trading system **Requirements**: - Implement multiple ML models for price prediction - Create backtesting framework with realistic constraints - Build performance analytics dashboard - Address risk management and regulatory considerations **Deliverables**: - Trading strategy implementation - Backtesting results and analysis - Risk assessment framework - Performance comparison with benchmarks #### Option 2: Financial NLP Application **Objective**: Build sentiment analysis system for financial markets **Requirements**: - Multi-source sentiment analysis (news, social media, earnings calls) - Correlation analysis with market movements - Real-time monitoring dashboard - Model validation and performance evaluation **Deliverables**: - NLP pipeline implementation - Sentiment-price correlation analysis - Interactive monitoring dashboard - Business case and ROI analysis #### Option 3: Robo-Advisory Platform **Objective**: Create automated portfolio management system **Requirements**: - Risk profiling questionnaire and algorithm - Modern Portfolio Theory implementation - Rebalancing and tax optimization - Client reporting and communication system **Deliverables**: - Complete robo-advisory system - Portfolio optimization algorithms - Client interface and reporting - Regulatory compliance framework #### Option 4: Credit Risk AI System **Objective**: Develop ML-based credit assessment platform **Requirements**: - Advanced feature engineering with alternative data - Multiple ML model comparison - Bias detection and fairness algorithms - Model governance and monitoring framework **Deliverables**: - Credit scoring model implementation - Bias audit and mitigation strategies - Model monitoring dashboard - Regulatory compliance documentation ### Assessment Criteria #### Technical Implementation (40%) - **Code Quality**: Clean, well-documented, reproducible code - **Algorithm Selection**: Appropriate choice of methods and techniques - **Data Handling**: Proper preprocessing and validation - **Performance**: Model accuracy and computational efficiency #### Business Application (25%) - **Problem Definition**: Clear business case and objectives - **Implementation Strategy**: Realistic deployment considerations - **Risk Assessment**: Identification and mitigation of key risks - **Value Proposition**: Clear articulation of business benefits #### Critical Analysis (20%) - **Literature Review**: Integration of relevant academic and industry sources - **Methodology Justification**: Rationale for chosen approaches - **Limitations Discussion**: Honest assessment of model constraints - **Alternative Approaches**: Consideration of other possible solutions #### Communication (15%) - **Report Quality**: Professional writing and presentation - **Visualization**: Effective charts, graphs, and dashboards - **Documentation**: Clear technical documentation - **Referencing**: Proper academic citation format --- ## Assessment Support ### Preparation Resources - **Sample Projects**: Available on course GitHub - **Code Templates**: Starter notebooks for each project type - **Rubric Details**: Comprehensive marking criteria - **Past Examples**: Anonymized high-quality submissions ### Getting Help - **Office Hours**: Weekly consultation sessions - **Peer Review**: Structured feedback sessions - **Technical Support**: Lab assistants for coding issues - **Writing Support**: University writing center resources ### Submission Guidelines - **File Naming**: `SurnameFirstNameBNumber_ProjectTitle` - **Code Repository**: GitHub with clear README - **Report Format**: Word document via Turnitin - **Late Penalties**: As per university policy --- ## Academic Integrity ### Permitted Collaboration - **Discussion**: General concepts and approaches - **Code Review**: Peer feedback on implementation - **Resource Sharing**: Publicly available datasets and libraries ### Prohibited Activities - **Code Copying**: Submitting someone else's implementation - **Report Plagiarism**: Copying text from any source without attribution - **Data Fabrication**: Creating fake results or datasets ### AI Tools Policy - **ChatGPT/Copilot**: Permitted for learning and debugging - **Code Generation**: Must be acknowledged and understood - **Report Writing**: AI assistance must be disclosed - **Final Responsibility**: You must understand and defend all submitted work --- ## Grade Boundaries | Classification | Percentage Range | Description | |----------------|------------------|-------------| | **First Class** | 70-100% | Outstanding work demonstrating innovation and mastery | | **Upper Second** | 60-69% | Good quality work with solid understanding | | **Lower Second** | 50-59% | Acceptable work meeting basic requirements | | **Third Class** | 40-49% | Adequate work with limited demonstration of skills | | **Fail** | 0-39% | Insufficient demonstration of learning outcomes | --- *For detailed rubrics and additional support materials, see the [Course Handbook](../FIN510_Handbook_2025.html).*