Abstract Sciaenids are a kind of economic oceanic fish with the most dynamic vocal activity. It is important to find out what type of sound a soniferous fish can make. Therefore, I used this acoustic character as a tool to study the distribution of sciaenids in order to manage their fisheries. In the previous researches, scientist found that there were eight types of sounds (A~H type) appearing on the coastal areas of Yunlin, and the frequency range of the H-type sound could reach to 8000Hz. This sound was suspected to be made by Otolithes ruber. Connaughton (1994) and Sprague (2000) mentioned that the differences of length, weight, and tension of the sonic muscle and the amount of protein and glycogen affect the types of sounds emitted. To reveal the producer of high-frequency sound, this research analyzed and compared the length, width, thickness and somatic of sonic muscle and the amount of protein and glycogen in seven sciaenid species occuring on the west coast of Taiwan (Johnius tingi, Johnius sina, Johnius amblycephalus, Johnius amblycephalus, Pennahia argentata, Chrysochir aureus, and Otolithes ruber), then to find if the differences of physic parameter and energy supplication are the major factors making O. ruber produce high frequency of sound. In this research, I found that the length of O. ruber¡¦s sonic muscle was shorter than that other of sciaenids. However the weight of sonic muscle did not significantly differ although it was still the lightest one (2.33¡Ó1.00g). The amount of protein (20.37¡Ó0.67¢M) and glycogen (0.33¡Ó0.11¢M) in O. ruber was higher than that of other species and the cross section of sonic muscle fiber was smaller. These characteristics are suitable for O. ruber to make a high-frequency sound. Because of the sound is believed to be made by rubber, so passive sonar was applied to investigate the distribution of this sound type on the estuaries along the west coast of Taiwan (Tam-Shui River, Tou-Chien River, Ta-Chia River, Cho-sui River, Zeng-Wen River, and Kao-Ping River) to represent distribution of O. ruber and its seasonal change of vocal activity. Acoustic activity reached it peak in spring then decreased through summer, autumn, and winter. More sounds were found in the estuaries south of the Cho-sui River (including Cho-sui, Zeng-Wen, and Kao-Ping River) than those north of this river (including Ta-Chia, Tou-Chien, and Tam-Shui River).

Increased strain on incumbent wastewater distribution networks originating from population increases as well as climate change calls for enhanced resource utilization. Accurately being able to predict network interconnectivity is vital within the wastewater industry to enable operational management strategies that optimizes the performance of the wastewater system. In this thesis, an evaluation of the network interconnectivity prediction performance of two machine learning models, the multilayer perceptron (MLP) and the support vector machine (SVM), utilizing supervisory control and dataacquisition (SCADA) system data for a wastewater system is presented. Results of the thesis imply that the MLP achieves the best predictions of the network interconnectivity. The thesis concludes that the MLP is the superior model and that the highest achievable network interconnectivity accuracy is 56% which is attained by the MLP model. / Den ökade påfrestningen på nuvarande avloppsnät till följd av befolkningstillväxt och klimatförändringar medför att det finns behov för optimerad resursförbrukning. Att korrekt kunna predicera ett avloppsnät är önskvärt då det möjliggör för effektivitetshöjande operativ förvaltning av avloppssystemet. I denna avhandling evalueras hur väl två maskininlärningsmodeller kan predicera nätverketssammankoppling med data från ett system för övervakning och kontroll av data (SCADA) genererat av ett avloppsnätverk. De två modellerna som testas är en multilagersperceptron (MLP) och en stödvektormaskin (SVM). Resultaten av avhandlingen visar på att MLP modellen uppnår den bästa prediktionen av nätverketssammankoppling. Avhandlingen konkluderar att MLP modellen är den bästa modellen för att predicera nätverkets sammankoppling samt att den högsta nåbara korrektheten var 56% vilket uppnåddes av MLP modellen.