TOPIC 1: miRNA detection by using surface plasmon resonance based sensor with different signal amplification technique

miRNAs are a class of endogenous, single-stranded, non-coding RNA molecules approximately 18–24 nucleotides in length. Numerous studies have shown that they are present in nearly all human cells as well as in various model organisms. miRNAs play critical roles in cell development, proliferation, differentiation, metabolism, apoptosis, and tumorigenesis. They also function as post-transcriptional regulators of gene expression in animals, plants, and viruses. Many experimental findings indicate that normal miRNA expression is essential for maintaining proper growth, reproduction, differentiation, and apoptosis in living organisms. In contrast, abnormal miRNA expression is closely associated with the onset of human diseases, genetic disorders, and alterations in immune system function. Therefore, the analysis and detection of miRNAs are highly important for biological research, disease diagnosis, and therapeutic development. However, miRNA detection remains challenging due to their short sequences, high similarity among homologous targets, low abundance, and the lack of universal features for selective amplification. This project aims to develop a novel surface plasmon resonance–based sensor for the detection of various miRNAs using different signal amplification strategies. We are seeking students who are interested in biosensors, biosample preparation, experimental design and execution, and large-scale data analysis.

TOPIC 2: Large scale rice yield forcasting by using machine learning and deep learning models

Rice is the primary food source for more than 50% of the global population and plays a significant role in global food security. Approximately 85% of rice is cultivated in Asia, including Taiwan. Therefore, accurate rice yield forecasting is essential for ensuring food security and supporting effective agricultural planning.  In this proposal, various machine learning and deep learning models are employed to predict large-scale rice yields in Taiwan based on meteorological factors.  Different machine learning models, such as Extreme Gradient Boosting (XGB), Random Forest (RF), Gradient Boosting Machine (GBM), Ridge Regression, Linear Regression, and Support Vector Machine (SVM)—are applied to rice yield prediction across Taiwan. In addition, different deep learning models, such as Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), and CNN-LSTM are used for rice yield prediction across Taiwan. The goal of this project is to build a smart prediction system for rice yield in Taiwan based on meteorological factors by using machine learning and deep learning models.  The results of this project provide several benefits for Taiwan, including helping farmers improve management practices, enabling traders and insurers to adjust pricing, assisting suppliers in managing inventories, and supporting government agencies in balancing rice imports and exports to ensure food security. In particular, authorities can rapidly mobilize resources in anticipation of potential yield shortfalls.