全國中小學科展

二等獎

Multiple Time-step Predictive Models for Hurricanes in the North Atlantic Basin Based on Machine Learning Algorithms

The cost of damage caused by hurricanes in 2017 is estimated to be over 200 billion dollars. Quick and accurate prediction of the path of a hurricane and its strength would be very valuable in alleviating these losses. Machine learning based prediction models, in contrast to models based on physics, have been developed successfully in many problem domains. A machine learning system infers the modeling function from a training dataset. This project developed machine learning based prediction models to forecast the path and strength of hurricanes in the North Atlantic basin. Feature analysis was performed on the HURDAT2 dataset, which contains paths and strengths of past hurricanes. Artificial Neural Networks (ANNs) and Generalized Linear Model (GLM) approaches such as Tikhonov regularization were investigated to develop nine hurricane prediction models. Prediction accuracy of these models was compared using a testing dataset, disjoint from the training dataset. The coefficient of determination and the mean squared error were used as performance metrics. Post-processing metrics, such as geodesic error in path prediction and the mean wind speed error, were also used to compare different models. TLS linear regression model performed the best of out the nine models for one and two time steps, while the ANNs made more accurate predictions for longer periods. All models predicted location and strength with greater than .95 coefficient of determination for up to two days. My models predicted hurricane path in under a second with accuracy comparable to that of current models.

Co-evolution of transcription factors and their binding sites in DNA

The synthesis of RNA using DNA (transcription) can be regulated by special proteins - transcription factors (TF) by binding to specific DNA regions - binding sites (BSTF). The purpose of the project is building a phylogenetic tree of orthologous groups of the each studying TF subfamilies, compare it with the tree of the corresponding BSTF motifs belonging to one orthologous group, find some common parts.

Algae Meets Fungi: Microalgae-Fungi Co-Pelletization for Biofuel Production

Microalgae-fungi biofuel has significantly less CO2 emissions than fossil fuels, making it much more environmentally friendly. As well, unlike traditional biofuel, microalgae-fungi does not require large masses of agricultural land for production. Thus, microalgae-fungi is an optimal option for biofuel production. This is a cost-effective renewable energy source that can be used in place of regular gas in cars and other means of transportation. By determining the most effective fungi for biofuel production, the threat of the impending environmental damage from pollution can be diminished. This novel experiment determines which fungi: Aspergillus niger, Rhizopus stolonifer or Saccharomyces cerevisiae, is the most effective bioflocculant in the microalgae-fungi co-pelletization process for biofuel production. We hypothesize that when paired with the microalgae Chlorella vulgaris, Rhizopus stolonifer will be the most effective. It has a high lipid content which could enhance the overall production of biofuel. Furthermore, its negative charge will aid with attracting and neutralizing the C. vulgaris colloidal particles resulting in an easier and more efficient removal of microalgae particles. Through the process of bioflocculation, pelletization, esterification and transesterification, the most effective fungi paired with C. vulgaris was determined. This experiment was carried out thoroughly and precisely resulting in a cost-effective solution for the world's current pollution crisis.

以時頻分析方法量化自主神經活性

許多疾病與自主神經活性息息相關,例如睡眠呼吸中止症。現有的分析方法,例如頻率分析法,將心律變異訊號(HRV)訊號假定為平穩過程,使得我們只能得到這段訊號內的平均值,無法知道其即時資訊。本研究成功以短時距時頻分析轉換(Short-time Fourier transform)配合重新分佈方法(reassignment method)來分析自主神經系統的活性。藉由觀察隨時間變化的頻率,我們可以得知即時的交感神經與副交感神經活動。本研究可延伸應用於睡眠呼吸中止症的診斷。

液滴凝成奇異形狀的研究

乾冰上方放一銅片,在銅片上方滴一液滴後,液體由下向上凝固成冰,最後在頂端形成圓錐形、圓弧形或煙囪形等三種不同奇異形狀;這些奇異形狀的形成取決於液體的分子力以及液體分子和冰之間的附著力大小,亦即由接觸角所決定。利用微積分可推導出接觸角、頂角、以及冰液相對密度間的關係式。 水在凝固過程中,利用Hele Shaw Cell量得接觸角為90°,水滴頂端凝成頂角65°的圓錐形。乙醇水溶液和葡萄糖水溶液的接觸角均隨溶液的濃度的增加而遞減且均小於90°,液滴頂端都凝成圓錐形。氯化鈉、氯化鉀、氯化鎂、氯化銨等四種不同濃度(0.75m~3m)的氯化物水溶液,高濃度的液滴頂端形成圓弧形,低濃度者形成煙囪形,濃度居中者形成圓錐形。2m的氯化銨水溶液液滴頂部先凝成煙囪,煙囪內再向上長出圓錐形。由推導出的關係式可解釋這些液滴凝成的奇異的現象。

神農菌、嚐百草-利用基因轉殖的大腸桿菌偵測中草藥內的重金屬

中藥一直存有重金屬汙染的疑慮,而目前檢驗這些可疑重金屬的儀器與技術皆需耗費大量金錢與時間,因此研發出簡便快速的偵測工具極為迫切。本研究目的是利用基因轉殖的大腸桿菌,偵測出中草藥內的重金屬。 實驗結果顯示,利用含有銅離子基因轉殖的大腸桿菌,偵測7種常見的中藥材浸膏,除了黃岑之外,其他6種中藥材內可偵測到銅離子,並且具有定量的螢光表現。其中,研究發現由於黃岑本身會有吸附銅離子的現象,因此利用螯合劑EDTA來解決此問題。結果顯示,螯合劑不會影響大腸桿菌的正常代謝,但能成功地將銅離子搶出並誘導細菌產生螢光。最後利用螢光顯微鏡觀察,發現深色的中藥並不會影響大腸桿菌的螢光表現。 未來期望可將本研究初步成果,做更進一步的研究,除了能合成更多不同基因的大腸桿菌,以偵測不同種類的重金屬外,還製成便宜且方便使用的重金屬生物感測器產品,方便民眾檢驗手中的中草藥材,確保所使用的中草藥之安全性。

波濤洶湧的玻璃杯—玻璃杯共振之研究

一個裝有水的玻璃杯,用喇叭撥出大聲的聲音,便能觀察到共振的現象發生,玻璃杯將會不停的震動,甚至導致破裂。本研究將會藉由傅立葉轉換(Fourier transform)來分析玻璃杯的振動的本徵頻率,並進一步探討頻譜的峰值。另外我們也研究了不同水位的高低以及液體密度對於共振頻率的影響,並以能量守恆的觀點進一步推導出玻璃杯的震動方程式。最後我們也近一步討論higher harmonic,並且發現non-linear dispersion relation的現象,有別於我們對於f∝n的觀念。

命中「助」定-間接互助模型的探討

每個人都有需要幫助的時候,當你遇到一個需要幫助的人,你會如何反應呢?有些人總是樂於助人,有些人選擇獨善其身,有些人則是會先觀察對方是什麼樣的人再做決定。這些不同的決定會交織成出什麼樣的故事呢? 我們假設社會上有三種人: 總是願意幫助別人的Cooperators、永遠不幫助別人的Defectors、依據人們過往的行為來決定是否幫助對方的Discriminators;由於cooperator和defector的行為是固定的,顯然discriminator的助人行為會有決定性的影響,因此我們從學者Berger的論文出發,修改discriminator決定是否幫助他人的判斷準則,架構了兩種間接互助模型並與Berger(2011)的工作做比較,觀察這些改變的影響,計算三種人的比例與彼此幫助率的關係。 Discriminator遇到不同的人會有不同的決定,隨著時間的推進,Berger在論文中已證明了他們的行為會趨於一致,並提出一種演化機制,探討三種族群之間的流動。那麼在我們提出的兩種模型中,discriminator的行為是不是也會趨於一致呢?因此我們證明了discriminator助人行為的收斂,也提出了一個新的演化機制,試圖用不同的觀點詮釋三種族群間的流動。 在我們的互助模型中,當使用Berger(2011)的演化機制時,演化行為只受discriminator佔全體比例的大小影響;然而若使用我們提出的演化機制,不論discriminator的比例為何,演化行為只會由 cooperator與defector 的比值決定。如此,我們便能刻劃出:可以使模型中所有的人至終演化成為 cooperators,理想中大同世界的範圍。

小花蔓澤蘭活性成分Dihydromilanolide誘導胃癌細胞凋亡與自噬性死亡

外來入侵種小花蔓澤蘭繁殖速度太快,使台灣本土生環境及多樣性受到破壞。我們研究發現,小花蔓澤蘭葉萃取物會誘導人類胃癌(AGS)細胞毒性;以HPLC分析及分離出小花蔓澤蘭葉萃取物的活性成分Dihydromilanolide (DHK),發現DHK會毒殺胞胃癌、卵巢癌、乳癌與血癌細胞,其中以胃癌(AGS)細胞毒殺性最強。此外,抗氧化劑N-acetylcysteine可減緩小花蔓澤蘭葉萃取物及DHK對胃癌細胞毒殺性,推測是透過活性氧化物(ROS)來毒殺胃癌細胞。我們亦發現,DHK可與抗癌藥物(Doxorubicin、Cisplatin或Paclitaxel)對胃癌細胞產生協同作用。DHK作用胃癌細胞會誘導Caspase-3增加、PARP蛋白裂解、促凋亡Bax增加及抑凋亡Bcl-2減少;產生酸性囊泡(AVOs)、促自噬LC3-II及Beclin-1蛋白增加,且加入自噬抑制劑 3-MA可保護DHK誘導胃癌細胞死亡。推論 DHK可誘發胃癌細胞凋亡(Apoptosis)及自噬性死亡(Autophagy)。總結,小花蔓澤蘭與活性成分Dihydromilanolide (DHK) 具抗胃癌功效且與抗癌藥物產生協同作用,可開發成抗癌的藥品或保健食品。

當機立「斷」—— 浮萍自裂脫險的機制與生態意義

浮萍在逆境下葉狀體會有分離的現象,本研究證實:浮萍透過葉狀體分離,增加逃離逆境的機率,提升族群生存率。此分離機制受到過氧化物質(H2O2)的調控,逆境下,浮萍母葉節處(node)的H2O2含量上升並誘導細胞死亡,進而造成連接構造斷裂,另能透過乙烯途徑活化纖維素分解酶使葉狀體分離。我們也發現青萍及紫萍具不同生存策略:青萍對H2O2的高敏感度使其能在逆境下快速分離,進而降低其葉狀體間的內聚力,更容易藉由風吹或水流加速逃離逆境;紫萍則對H2O2較不敏感且內聚力大,以較大的單一個體及對逆境的高耐受性來渡過危機。蛋白質含量極高的浮萍是蛋白質補給品的好原料,期待分離機制的深入研究能應用在浮萍種植上,使其快速分離提升產量,應對將到來的糧食危機。