Browsing by Author "Li, Lihong"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Effects of Source and Level of Trace Minerals on Performance, Mineral Excretion, Intestine and Bone Development, and Immune Response in Commercial Turkeys
    Li, Lihong (Virginia Tech, 2009-01-23)
    To compare the effect of a standard commercial trace mineral dietary program to low levels of organic minerals on turkey performance, mineral excretion, bone strength, and carcass yield, day-old Hybrid poults (n=1,224) were randomly distributed to one of four treatments with 9 replicates. Experimental treatments consisted of: standard inorganic (SI) with a commercial supplementation program (Mn, Zn, Cu, Se), reduced inorganic (RI) with 10% level of SI, and two organic regimens of Bioplex®/Sel-Plex® (at the same level of RI during period 1 and 2 and at 2/3 of RI for period 3, 4, 5, and 6, or at the same level of RI for entire trial). Body weight (BW), body weight gain (BWG), feed conversion ratio (FCR), and feed intake (FI) were evaluated and fresh excreta were collected at d 28, 49, 70, 84, 105 and 133. Tibias and femurs were collected at d 49, 84 and 133. Trace mineral concentration in litter and carcass yield were determined at d 133. Overall, there was no significant effect on BW, cumulative BWG, FCR, or FI due to treatments (P < 0.05). The contents of Mn and Zn in excreta and litter were significantly reduced (P < 0.05) in Bioplex®/Sel-Plex® or RI diet compared to SI during the study. Cu excretion was significantly reduced at d 84 and 133. Tibias from the SI treatment had increased bone strength at d 49. Carcass yield at processing was significantly improved (P < 0.05) by feeding Bioplex®/Sel-Plex® treatments compared to the SI diet. To investigate the effect of organic or inorganic Zn combined with other trace minerals on turkey performance, immune response, and intestinal development, a 2 by 4 factorial design was utilized with coccidia vaccinated and non-vaccinated and 4 dietary treatments varying in level and source of Zn with Mn, Cu, and Se. A total of 2,376 day-old Hybrid turkeys were assigned to one of the combinations with 9 replicates of each. Dietary treatments consisted of: 1) standard inorganic (SI), Zn (150 ppm) with Mn (165 ppm), Cu (10 ppm), and Se (0.2 ppm); 2) reduced inorganic (RI), Zn, Mn, and Cu at 10% of SI, and Se at 0.2ppm; 3) organic 1 (O1), at the same level of RI; 4) organic 2 (O2), Zn (30 ppm) with the same level of Mn, Cu, and Se as O1. Body weight, BWG, FI and FCR were determined weekly. Bursa, thymus, and spleen were weighed, and duodenum and jejunum were collected at d 7, 14, 28, and 42. Peripheral blood was collected for T-lymphocyte populations on d 21, 28, and 42. Cumulative FI was influenced by vaccination (P=0.003). Cumulative BWG and BW were significantly decreased by vaccination except on d 14. Cumulative BWG increased in poults fed RI compared with those fed O2 (P=0.03). Poults fed O2 had significantly decreased BW when compared with RI after d 28. Cumulative FCR was not affected by diet and vaccination. Vaccination increased spleen weight on d 7 and thymus weight on d 42 (P < 0.05). The birds fed O2 had increased thymus weight when compared with those fed SI at d 7 (P < 0.05). The vaccinated poults had higher numbers of CD4+ T-cells than non-vaccinated birds on d 28 and d 42 (P < 0.05), and an interaction between diet and vaccination was observed (P < 0.05). Compared to non-vaccinated poults, CD4+/CD8+ ratio was significantly increased in vaccinated poults on d 42 (P = 0.0475). The villus height in vaccinated birds was significantly increased in the jejunum (P = 0.0012), but diets did not affect intestinal morphology. In summary, using low levels of organic or inorganic trace minerals is adequate to maintain turkey performance and immune response and decreased trace minerals excretion.
  • Thumbnail Image
    Machine Learning in the Bandit Setting: Algorithms, Evaluation, and Case Studies (CS Seminar Lecture Series)
    Li, Lihong (2012-02-10)
    Much of machine-learning research is about discovering patterns---building intelligent agents that learn to predict future accurately from historical data. While this paradigm has been extremely successful in numerous applications, complex real-world problems such as content recommendation on the Internet often require the agents to learn to act optimally through autonomous interaction with the world they live in, a problem known as reinforcement learning. Using a news recommendation module on Yahoo!'s front page as a running example, the majority of the talk focuses on the special case of contextual bandits that have gained substantial interests recently due to their broad applications. We will highlight a fundamental challenge known as the exploration/exploitation tradeoff, present a few newly developed algorithms with strong theoretical guarantees, and demonstrate their empirical effectiveness for personalizing content recommendation at Yahoo!. At the end of the talk, we will also summarize (briefly) our earlier work on provably data-efficient algorithms for more general reinforcement-learning problems modeled as Markov decision processes. Bio: Lihong Li is a Research Scientist in the Machine Learning group at Yahoo! Research. He obtained a PhD degree in Computer Science from Rutgers University, advised by Michael Littman. Before that, he obtained a MSc degree from the University of Alberta, advised by Vadim Bulitko and Russell Greiner, and BE from the Tsinghua University. In the summers of 2006-2008, he enjoyed interning at Google, Yahoo! Research, and AT&T Shannon Labs, respectively. His main research interests are in machine learning with interaction, including reinforcement learning, multi-armed bandits, online learning, active learning, and their numerous applications on the Internet. He is the winner of an ICML'08 Best Student Paper Award, a WSDM'11 Best Paper Award, and an AISTATS'11 Notable Paper Award. The Computer Science Seminar Lecture Series is a collection of weekly lectures about topics at the forefront of contemporary computer science research, given by speakers knowledgeable in their field of study. These speakers come from a variety of different technical and geographic backgrounds, with many of them traveling from other universities across the globe to come here and share their knowledge. These weekly lectures were recorded with an HD video camera, edited with Apple Final Cut Pro X, and outputted in such a way that the resulting .mp4 video files were economical to store and stream utilizing the university's limited bandwidth and disk space resources.