Enterprise AI Analysis
Predicting Visual Aesthetic Preferences in Tehran City Universities Campuses Using Machine Learning Techniques
Visual aesthetic preferences fundamentally shape the restorative potential of university landscapes and have a significant impact on student well-being and engagement. This study developed Ensemble Learning Models to predict students' aesthetic preferences for interactive rest spots and compared their accuracy with conventional individual models. The input dataset (18 features) was extracted from images of 100 student rest spots across four universities campuses in Tehran city: University of Tehran, Amirkabir University of Technology, Shahid Rajaee Teacher Training University, and Tarbiat Modares University. Based on the aesthetic preferences reported by 394 university students, the study employed Support Vector Regression (SVR), Random Forest (RF), Multilayer Perceptron (MLP), and their combinations of SVR-MLP and SVR-RF-MLP to predict the aesthetic quality on university campuses. The results show that Ensemble Learning Models outperform individual models in predicting students' aesthetic preferences, filling a key research gap. The individual models demonstrated varying levels of accuracy across the total dataset, with SVR (R2 = 0.824) performing the strongest, followed by MLP (R2 = 0.814) and RF (R2 = 0.761). Among all, the SVR-MLP ensemble learning model achieved the highest accuracy, with R² scores of 0.767 (test data), 0.850 (training data), and 0.828 (total dataset). Key design elements enhancing both aesthetic appeal and mental restoration included more trees, soft landscapes, waterscapes, and color diversity, coupled with minimal building and pathway presence. The Ensemble Learning Models provide a robust conceptual framework for architects, environmental designers, landscape architects, and campus planners to design attractive and restorative spaces aligned with students' visual preferences.
Quantified Impact of Ensemble Learning Models
The application of advanced machine learning significantly enhances our ability to predict complex human preferences for restorative environments.
0
Highest R² Score (Total Dataset)
0
University Students Surveyed
0
Environmental Features Analyzed
0
Tehran Universities Included
Deep Analysis & Enterprise Applications
Select a topic to dive deeper, then explore the specific findings from the research, rebuilt as interactive, enterprise-focused modules.
Enterprise Process Flow
Review Literature & Identify Features
→
Select Tehran Universities
→
Sample Restorative Spots
→
Photograph & Preprocess Images
→
Distribute & Score Questionnaire
→
Develop ML Algorithms (SVR, MLP, RF)
→
Run ML Algorithms & Sensitivity Analysis
→
Optimize Model with Sensitivity Analysis
Robust Data Acquisition & Model Development
This study employed a sequential mixed-methods approach for data acquisition, combining spatial analysis and on-site surveys to identify 100 student-preferred rest spots across four Tehran universities. Photographic documentation followed a standardized protocol to ensure visual consistency, capturing images at eye-level with controlled conditions. These images served as visual stimuli for an online survey, where 394 students rated aesthetic preferences on a 7-point scale. The extracted dataset comprised 18 environmental features, including architectural elements, landscape components, and facilities, theoretically grounded in Attention Restoration Theory (ART), Stress Reduction Theory (SRT), and Socially Restorative Urbanism (SRU). For predictive modeling, Support Vector Regression (SVR), Multilayer Perceptron (MLP), and Random Forest (RF) were implemented, alongside two ensemble learning models (SVR-MLP and SVR-MLP-RF). Data preprocessing involved random splitting into training (80%) and testing (20%) subsets, followed by feature standardization using Python's StandardScaler() to prevent data leakage.
0.828
Total R² Score for SVR-MLP Ensemble Model
| Model | Test R² | Train R² | Total R² | Test MAE | Total MAE |
|---|---|---|---|---|---|
| SVR | 0.7617 | 0.8474 | 0.8249 | 0.4097 | 0.2698 |
| MLP | 0.7592 | 0.8338 | 0.8142 | 0.4156 | 0.3226 |
| RF | 0.7039 | 0.7813 | 0.7610 | 0.4910 | 0.3803 |
| SVR-MLP Ensemble | 0.7672 | 0.8505 | 0.8286 | 0.4026 | 0.2887 |
| SVR-MLP-RF Ensemble | 0.7623 | 0.8412 | 0.8205 | 0.4031 | 0.3034 |
| Ensemble models, especially SVR-MLP, demonstrate superior predictive accuracy and robustness. | |||||
Key Influential Environmental Features
The sensitivity analysis, conducted using the SVR-MLP ensemble model, identified the top six influential features for student aesthetic preferences: tree cover, soft landscape, waterscape, color diversity, building presence, and pathway presence. Notably, building and pathway presence showed an inverse relationship with aesthetic preference, suggesting that less visual prominence of these elements contributes to higher aesthetic appeal. Conversely, natural elements like trees, soft landscapes, waterscapes, and color diversity significantly enhance perceived aesthetic quality. These findings align with established environmental psychology theories, underscoring the importance of natural elements for restorative campus environments.
Machine Learning Advances & Design Relevance
This study effectively bridges computational prediction with theoretical insights from environmental psychology, offering a practical framework for integrating human-centered visual evaluation into campus planning. The ensemble learning models, particularly SVR-MLP, demonstrated robust predictive accuracy for assessing aesthetic preferences, outperforming individual models and filling a key research gap in campus aesthetics research. Unlike previous studies relying on limited visual indicators or qualitative assessments, this work adopted a comprehensive, theory-driven feature selection approach. The findings confirm that ML models are powerful tools for understanding complex, subjective environmental preferences and can inform evidence-based design decisions.
Strategic Design Implications for University Campuses
This research provides a foundation for developing data-driven approaches to the design of interactive university spaces. Enhancing landscape aesthetics and mental restoration potential requires the integration of diverse natural elements across campus environments. For instance, a combination of deciduous and evergreen trees can diversify winter scenery and promote ecological resilience. Alongside tree diversity, planting colorful flowers and a variety of vegetation—combined with small fountains or larger water features—offers another effective way to enhance visual appeal and meet students’ aesthetic preferences. In addition to visual richness, physical comfort also plays a significant role in shaping students’ spatial experience. Redesigning campus pathways plays a critical role in improving user comfort. The selection of appropriate paving materials such as brick, wood, stone, or tile—combined with grass strips instead of concrete and asphalt—along with shading elements like vegetation or semi-transparent canopies, contributes to thermal comfort. Seating and interaction nodes aligned with the Social Restorative Urbanism (SRU) framework may further enhance spatial experience, particularly when integrated along pathway edges without obstructing pedestrian flow. These design strategies not only offer improved aesthetic quality but also reduce mental fatigue and increase users’ sense of spatial presence.
- Increase green spaces and waterscapes for restorative effects.
- Enhance color diversity with varied plant species.
- Minimize visual prominence of buildings and hard pathways.
- Implement ergonomically designed seating and shaded areas.
Empowering Evidence-Based Campus Design
This study successfully developed and validated Ensemble Learning Models to predict students' aesthetic preferences for university rest spots in Tehran, achieving high predictive performance (SVR-MLP Total R² = 0.828). The findings underscore the critical role of natural elements—trees, soft landscapes, waterscapes, and color diversity—in fostering restorative environments, while also highlighting the inverse relationship with prominent built structures and pathways. By integrating machine learning with environmental psychology theories (ART, SRT, SRU), this research offers a robust, data-driven framework for architects, environmental designers, and campus planners to create aesthetically pleasing and psychologically supportive spaces. Future research should explore multi-seasonal data, indoor spaces, cross-cultural validations, and automate feature extraction to further enhance scalability and generalizability.
Limitations & Future Directions
While providing significant insights, this study acknowledges limitations including relatively low student participation rates in the questionnaire and technical errors in online response storage. Methodologically, the use of two-dimensional images might restrict the complete representation of perceptual and emotional experiences, though controlled environmental simulations remain valuable. Future research is encouraged to explore multi-seasonal data, indoor rest spaces, and cross-cultural validations to generalize findings. Integrating biometric indicators (e.g., eye-tracking, EEG) and automating expert-based feature extraction through advanced image recognition or LiDAR-based analysis will further enhance prediction accuracy, capture implicit aesthetic responses, and improve scalability for evidence-based landscape design.
Calculate Your Potential ROI with AI
Understand the tangible benefits of integrating AI into your enterprise operations with our interactive ROI calculator.
Estimated Annual Savings
$0
Productive Hours Reclaimed Annually
0
Your AI Implementation Roadmap
A clear path to integrating advanced AI, designed for maximum impact and minimal disruption.
Phase 1: Discovery & Strategy
Comprehensive analysis of your current operations, identification of AI opportunities, and development of a tailored strategy.
Phase 2: Pilot & Validation
Deployment of a small-scale AI pilot project to validate models, gather initial data, and demonstrate tangible value.
Phase 3: Scaled Integration
Full-scale integration of AI solutions across relevant departments, ensuring seamless adoption and continuous optimization.
Phase 4: Performance & Growth
Ongoing monitoring, performance tuning, and identification of new AI-driven growth opportunities and innovations.
Ready to Transform Your Enterprise with AI?
Book a free consultation with our AI specialists to explore how these insights can be applied to your specific business challenges.
Ready to Get Started?
Book Your Free Consultation.
Let's Discuss Your AI Strategy!
Lets Discuss Your Needs
Select a Date
Sun
Mon
Tue
Wed
Thu
Fri
Sat