Federal Polytechnic Ilaro Journal of Pure And Applied Sciences

A COMPARATIVE ANALYSIS OF EXTREME GRADIENT BOOSTING AND SUPPORT VECTOR REGRESSION FOR MODELING BENCHMARK CRUDE OIL PRICES

Nurudeen O. ALABI,; Gabriel O. OJO — Mon, 30 Dec 2024 00:00:00 +0000

Organization of Petroleum Exporting Countries (OPEC) and non-OPEC supply, oil balance, oil demand by Organization for Economic Cooperation and Development (OECD) and non-OECD members, money market managers' long positions, US consumer price index and spot prices of crude oils like New York Mercantile Exchange West Texas Intermediate (NYMEX WTI), Intercontinental Exchange (ICE) Brent, OPEC Reference Basket (ORB), and other crude oils are basic elements driving the patterns seen in the market pricing of crude oils. Data between 2008 and 2022 were obtained for this study from OPEC Monthly Oil Market Reports. This research evaluates the performance of two machine learning models, Support Vector Regression (SVR) and Extreme Gradient Boosting (XGBoost), in predicting crude oil prices for three major benchmarks: OPEC Reference Basket (ORB), NYMEX WTI, and ICE Brent. Using key performance metrics such as Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and R-squared (R²), the study highlights the strengths and weaknesses of each model in both stable and volatile market conditions. SVR shows strong predictive accuracy, particularly for ICE Brent, but struggles with price volatility in the ORB and NYMEX WTI datasets. XGBoost is more robust in handling volatility and non-linear relationships. The findings have important economic implications for market participants, suggesting that while SVR is suited for stable pricing environments, XGBoost is better equipped to handle the unpredictability of more volatile markets.

COMPARATIVE ANALYSIS OF PHYTOCHEMICAL CONSTITUENTS, PROXIMATE COMPOSITION AND ENERGY VALUE OF Hibiscus sabdariffa L. LEAF AND PETAL EXTRACTS

Deborah Etooluwa KUMOYE; Ifelolu Adeseye REMI-ESAN — Mon, 30 Dec 2024 00:00:00 +0000

This study evaluates the phytochemical constituents, proximate composition, and energy values of methanolic, aqueous and ethanolic extracts from Hibiscus sabdariffa L. leaves and petals. Qualitative phytochemical analysis showed the presence of bioactive compounds such as steroids, alkaloids, terpenoids, and flavonoids, with petal extracts showing significantly higher amounts of bioactive compounds across all solvents. The quantitative analysis indicated that aqueous extracts had the highest concentration of flavonoids at 267 mg/100 g, while ethanolic petal extracts had the highest terpenoid content at 7.5mg/100g. The proximate analysis demonstrated that the leaves possessed a greater energy value of 280.36 ± 1.20 kcal/100 g and a carbohydrate content of 45.56 ± 0.89%. In contrast, the petals had a higher protein content of 11.42 ± 0.38% and crude fibre content of 37.49 ± 0.64%. These findings underscore the nutritional and medicinal benefits of H. sabdariffa. The leaves can provide dietary energy, whereas the petals may have potential applications in antioxidant therapies and functional foods. Future research should focus on their pharmacological applications and potential industrial uses.

STUDENT ACADEMIC PERFORMANCE PREDICTION SYSTEM USING ENSEMBLE ALGORITHM

EMMANUEL AYODELE; Victor O SODEINDE — Mon, 30 Dec 2024 00:00:00 +0000

Predicting student academic performance is crucial for enhancing educational outcomes and supporting timely interventions. There has been an increased interest in creating precise models for projecting student performance as a result of the introduction of machine learning techniques and the accessibility of large-scale educational data. However, the creation of predictive models in educational contexts is frequently hampered by the sensitive nature of student data and the requirement to uphold privacy and data security. This study develops a student Academic performance prediction system that leverages ensemble techniques, specifically focusing on Random Forest and other robust machine learning methods. By analyzing data such as demographic attributes, behavioral factors, and prior academic records, the system identifies patterns that influence final grades, categorizing students' performance levels. Categorical data is encoded, and a test-train split methodology is employed to assess the model's predictive accuracy. The Random Forest Regressor is particularly effective in capturing complex patterns by combining multiple decision trees, resulting in a high degree of accuracy and reduced overfitting. The model’s effectiveness is measured by mean squared error (MSE) scores, indicating its ability to deliver precise predictions. Additionally, the system stores trained models and encoding schemes, facilitating real-time usage and scalability for larger datasets. This prediction system offers educators actionable insights for academic support, enhances individualized student guidance, and enables targeted educational strategies. Overall, the application of ensemble techniques in student performance prediction presents a valuable tool for data-driven decision-making in educational settings.

DEVELOPMENT OF A CRUCIBLE FURNACE FIRED WITH LIQUEFIED PETROLEUM GAS - BUTANE GAS

Enemona O. SANNI; Vincent B. ONIGBARA — Mon, 30 Dec 2024 00:00:00 +0000

This research work focuses on the development of a crucible furnace that runs on liquefied petroleum gas (LPG) LPG-butane gas. The refractory material (aluminosilicate) used for the construction of the furnace wall was harvested from Ilaro, Ogun State, Clay deposit. Other materials used not limited to crucible pot, 6 mm mild steel angle iron, household gas cylinder, 60 mm rubber hose, and mild steel with a thickness of 5 mm were sourced from a local market in Lagos. The performance evaluation of the constructed crucible furnace was ascertained by charging aluminium scrap and other ferrous metal into it with complete melting occurring at a furnace temperature of 750^oC

DESIGN AND CONSTRUCTION OF THERMOELECTRIC GENERATOR USING PARABOLIC TROUGH COLLECTOR

Enemona O. SANNI; Vincent B. ONIGBARA — Mon, 30 Dec 2024 00:00:00 +0000

attention to the need to reduce carbon emissions and balance the energy supply and demand. Thomas Johann Seebeck discovered in 1821 that electricity can be generated by a temperature gradient that forms between two different conductors. The diffusion of charge carriers is caused by heat flow, which is the fundamental mechanism underlying the thermoelectric effect in conducting materials. In turn, the movement of charge carriers between the hot and cool areas produces a voltage differential. A thermoelectric generator (TEG), or a Seebeck generator, is a solid-state device that uses the Seebeck effect to directly convert heat flow, or temperature differential, into electrical energy. This experimentation on the design and construction of a 1kW thermoelectric generator using solar energy with a parabolic trough collector is composed of a parabolic trough collector framed from a plain mirror with an aperture area of m 0.27m2 that was used to concentrate sunlight onto a copper receiver plate with an area of 10 × 10 cm². A power of 0.9kW was achieved by connecting 7 numbers of thermoelectric modules (TEM) generating 2.4V and 469 mA at an average temperature gradient of 60^oC.