Summary of the BOSO algorithm. by Luis V. Valcárcel (12700318)

An Ecological Benchmark of Photo Editing Software: A Comparative Analysis of Local vs. Cloud Workflows by Pierre-Alexis DELAROCHE (22092572)

“…Reproducibility Framework Container Orchestration # Kubernetes deployment manifest for reproducible environment apiVersion: apps/v1 kind: Deployment metadata: name: energy-benchmark-pod spec: replicas: 1 selector: matchLabels: app: benchmark-runner template: metadata: labels: app: benchmark-runner spec: nodeSelector: hardware.profile: "high-performance" containers: - name: benchmark-container image: albumforge/energy-benchmark:v2.1.3 resources: requests: cpu: "8000m" memory: "16Gi" nvidia.com/gpu: 1 limits: cpu: "16000m" memory: "32Gi" env: - name: MEASUREMENT_PRECISION value: "high" - name: POWER_SAMPLING_RATE value: "1000" # 1kHz sampling Dependency Management FROM ubuntu:22.04-cuda11.8-devel RUN apt-get update && apt-get install -y \ perf-tools \ powertop \ intel-gpu-tools \ nvidia-smi \ cpupower \ msr-tools \ && rm -rf /var/lib/apt/lists/* COPY requirements.txt /opt/ RUN pip install -r /opt/requirements.txt Usage Examples and API Documentation Python Data Analysis Interface import pandas as pd import numpy as np from scipy import stats import matplotlib.pyplot as plt import seaborn as sns # Load dataset with optimized dtypes for memory efficiency df = pd.read_csv('ecological_benchmark_dataset.csv', dtype={'hardware_config': 'category', 'test_type': 'category'}) # Compute energy efficiency metrics df['energy_per_photo'] = df['energy_consumption_kwh'] / df['photo_count'] df['co2_per_gigabyte'] = df['co2_equivalent_g'] / df['total_volume_gb'] # Statistical analysis with confidence intervals local_energy = df[df['test_type'] == 'local_processing']['energy_consumption_kwh'] cloud_energy = df[df['test_type'] == 'cloud_processing']['energy_consumption_kwh'] t_stat, p_value = stats.ttest_ind(local_energy, cloud_energy) effect_size = (cloud_energy.mean() - local_energy.mean()) / np.sqrt((cloud_energy.var() + local_energy.var()) / 2) print(f"Statistical significance: p = {p_value:.2e}") print(f"Cohen's d effect size: {effect_size:.3f}") R Statistical Computing Environment library(tidyverse) library(lme4) # Linear mixed-effects models library(ggplot2) library(corrplot) # Load and preprocess data df <- read_csv("ecological_benchmark_dataset.csv") %>% mutate( test_type = factor(test_type), hardware_config = factor(hardware_config), log_energy = log(energy_consumption_kwh), efficiency_ratio = energy_consumption_kwh / processing_time_sec ) # Mixed-effects regression model accounting for hardware heterogeneity model <- lmer(log_energy ~ test_type + log(photo_count) + (1|hardware_config), data = df) # Extract model coefficients with confidence intervals summary(model) confint(model, method = "Wald") Advanced Analytics and Machine Learning Integration Predictive Modeling Framework from sklearn.ensemble import RandomForestRegressor, GradientBoostingRegressor from sklearn.model_selection import cross_val_score, GridSearchCV from sklearn.preprocessing import StandardScaler, LabelEncoder from sklearn.metrics import mean_absolute_error, r2_score # Feature engineering pipeline def create_feature_matrix(df): features = df[['photo_count', 'avg_file_size_mb', 'total_volume_gb']].copy() # Polynomial features for capturing non-linear relationships features['photo_count_squared'] = features['photo_count'] ** 2 features['size_volume_interaction'] = features['avg_file_size_mb'] * features['total_volume_gb'] # Hardware configuration encoding le = LabelEncoder() features['hardware_encoded'] = le.fit_transform(df['hardware_config']) return features # Energy consumption prediction model X = create_feature_matrix(df) y = df['energy_consumption_kwh'] # Hyperparameter optimization param_grid = { 'n_estimators': [100, 200, 500], 'max_depth': [10, 20, None], 'min_samples_split': [2, 5, 10], 'min_samples_leaf': [1, 2, 4] } rf_model = RandomForestRegressor(random_state=42) grid_search = GridSearchCV(rf_model, param_grid, cv=5, scoring='neg_mean_absolute_error') grid_search.fit(X, y) print(f"Best cross-validation score: {-grid_search.best_score_:.6f}") print(f"Optimal hyperparameters: {grid_search.best_params_}") Carbon Footprint Calculation Methodology Emission Factor Coefficients Carbon intensity calculations employ region-specific emission factors from the International Energy Agency (IEA) database: EMISSION_FACTORS = { 'EU_AVERAGE': 0.276, # kg CO₂/kWh (European Union average 2024) 'FRANCE': 0.057, # kg CO₂/kWh (Nuclear-dominant grid) 'GERMANY': 0.485, # kg CO₂/kWh (Coal transition period) 'NORWAY': 0.013, # kg CO₂/kWh (Hydroelectric dominant) 'GLOBAL_AVERAGE': 0.475 # kg CO₂/kWh (Global weighted average) } def calculate_carbon_footprint(energy_kwh: float, region: str = 'EU_AVERAGE') -> float: """ Calculate CO₂ equivalent emissions using lifecycle assessment methodology Args: energy_kwh: Energy consumption in kilowatt-hours region: Geographic region for emission factor selection Returns: CO₂ equivalent emissions in grams """ emission_factor = EMISSION_FACTORS.get(region, EMISSION_FACTORS['GLOBAL_AVERAGE']) co2_kg = energy_kwh * emission_factor return co2_kg * 1000 # Convert to grams Citation and Attribution This dataset is released under Creative Commons Attribution 4.0 International (CC BY 4.0) license. …”

<b>AI for imaging plant stress in invasive species </b>(dataset from the article https://doi.org/10.1093/aob/mcaf043) by Erola Fenollosa (20977421)

“…The described extracted features were used to predict leaf betalain content (µg per FW) using multiple machine learning regression algorithms (Linear regression, Ridge regression, Gradient boosting, Decision tree, Random forest and Support vector machine) using the <i>Scikit-learn</i> 1.2.1 library in Python (v.3.10.1) (list of hyperparameters used is given in <a href="#sup1" target="_blank">Supplementary Data S5</a>). …”