Search Results for « ReCLive »

Social media, professional sports, and video games are driving rapid growth in live video streaming, on platforms such as Twitch and YouTube Live. Live streaming experience is very susceptible to short-time-scale network congestion since client playback buffers are often no more than a few seconds. Unfortunately, identifying such streams and measuring their QoE for network management is challenging, since content providers largely use the same delivery infrastructure for live and video-on-demand (VoD) streaming, and packet inspection techniques (including SNI/DNS query monitoring) cannot always distinguish between the two. In this paper, we design, build, and deploy ReCLive: a machine learning method for live video detection and QoE measurement based on network-level behavioral characteristics. Our contributions are four-fold: (1) We analyze about 23,000 video streams from Twitch and YouTube, and identify key features in their traffic profile that differentiate live and on-demand streaming. We release our traffic traces as open data to the public; (2) We develop an LSTM-based binary classifier model that distinguishes live from on-demand streams in real-time with over 95% accuracy across providers; (3) We develop a method that estimates QoE metrics of live streaming flows in terms of resolution and buffer stall events with overall accuracies of 93% and 90%, respectively; and (4) Finally, we prototype our solution, train it in the lab, and deploy it in a live ISP network serving more than 7,000 subscribers. Our method provides ISPs with fine-grained visibility into live video streams, enabling them to measure and improve user experience.

BACKGROUND: The National Cancer Data Base (NCDB), a joint project of the American College of Surgeons Commission on Cancer and the American Cancer Society, is a cancer management and outcomes data base for health care organizations. It provides a comparative summary of patient care that is used by participating hospitals and communities for self-assessment. The most current (1995-1996) breast cancer data on patients from low income zip codes are described here. METHODS: Since 1989, eight Calls for Data have been issued, yielding a total of 191,714 reports of non-Hispanic white patients with breast cancer for the years analyzed, 1995-1996. A total of 1961 hospital cancer registries have participated in at least one of the Calls for Data. RESULTS: A diverse range of breast cancer cases was reported from a variety of geographic locations and medical care environments. There were general similarities in the treatment of patients from the different income groups; however, some differences were reported. Among patients from lower income zip codes, 60.7% were age 60 years or older, compared with 55.1% from other income zip code groups. The AJCC stage distribution was reported as less favorable for patients from low income zip codes than for other patients. The percentage of patients from low income zip codes diagnosed as Stage 0 or I was 51.2%, compared with 55.9% of patients from the other income zip codes. Of patients from lower income zip codes, 12.1% were reported to have Stage III or IV disease, compared with 10.0% of patients from other income zip codes. Patients from low income zip codes received less tissue-sparing surgery. Of patients from low income zip codes, 14.9% received partial mastectomy with or without radiation or systemic therapy, compared with 18.3% of patients from other income zip codes. The percentage of patients from low income zip codes who received a partial mastectomy with axillary lymph node dissection was 23.3% for patients from other income zip codes, the percentage was 30.5%. Conversely, 49.8% of patients from lower income zip codes received a modified radical mastectomy, compared with 40.5% of patients from other income zip codes. CONCLUSIONS: Further improvements in the early diagnosis and surgical treatment of low income patients can probably be achieved. Programmatic activities that further explain or reduce the apparent nonpreferred treatment of some low income patients should be encouraged.

Video streaming continues to be the largest service delivered on the internet. This includes gaming videos, delivered both on-demand and live, where gaming footage is usually accompanied by a video of the player overlaid on top of the gameplay - resulting in Picture-In-Picture (PiP) content. Currently, PiP content is usually combined into a single video before being delivered to the client via technologies such as HTTP Adaptive Streaming (HAS). In this study, we investigated the QoE importance of gameplay and player elements in PiP gaming videos by varying the video quality of these elements individually. We conducted a subjective study, testing nine quality permutations based on three quality levels across three pieces of content from different gaming genres, with 30 participants recruited using an ethical crowdsourcing platform. We found that gameplay was significantly more important in terms of overall QoE, while the player element made a difference in only a few cases.

BACKGROUND: The National Cancer Data Base (NCDB), a joint project of the Commission on Cancer of the American College of Surgeons and the American Cancer Society, is a cancer management and outcomes data base for health care organizations. It provides a comparative summary of patient care that is used by participating hospitals and communities for self-assessment. This article describes the most current (1995) data. METHODS: Since 1989, 7 calls for data have been issued, yielding a total of 5,558,389 cancer patient reports for the years 1985-1995. A total of 1849 hospital cancer registries have participated in at least 1 of the calls for data. RESULTS: One thousand one hundred and fourteen hospitals from 50 states and the District of Columbia reported 655,627 cases for the diagnosis year 1995. The hospitals represented a wide range of sizes (187 [16.8%] with 1000+ cases annually, 405 [36.4%] with 500-999 cases annually, 255 [22.9%] with 300-499 cases annually, 211 [18.9%] with 100-299 cases annually, and 56 [5%] with < 100 cases annually) and types (21 [1.9%] National Cancer Institute [NCI]-recognized cancer centers, 119 [10.7%] government hospitals, 102 [9.2%] teaching hospitals, 256 [23.0%] large community hospitals, 297 [26.7%] medium/small community hospitals, and 257 [23.1%] nongovernmental hospitals without approval status from the Commission on Cancer or NCI recognition). Remarkably similar distributions of cases by primary site and age were reported from each of six U.S. geographic regions. In addition, within each of these six regions, the cases were reported from a wide range of income strata and ethnicities. For several states, relatively few cancer cases were reported. For several examples of relatively rare patient and tumor groups, all reported cases between 1985-1995 included potentially useful quantities of patients in whom further study of such special groups was warranted. CONCLUSIONS: The authors conclude that the reported cases most likely are representative at the regional (but not state) level of cancer patients diagnosed and treated at U.S. hospitals with regard to types of cancer and ages of the patients. They conclude further that cancer reporting may be quite diverse within each region with regard to other known patient and reporting institution characteristics.

As modern networks evolve into complex systems to support next-generation applications with strict latency requirements, in-switch machine learning (ML) has emerged as a candidate technology for minimizing ML inference latency. Multiple solutions, mostly based on Decision Tree (DT) and Random Forest (RF) models, have been proposed in that regard for inference at packet level (PL) or flow level (FL) or simultaneously at PL and FL. Such heterogeneity in the inference target leads to the use of varying performance metrics for evaluating the solutions, rendering a fair comparison between them difficult. In this paper, we perform a comprehensive evaluation of 5 leading solutions for DT/RF-based in-switch inference. We replicate and deploy the solutions into a real-world testbed with Intel Tofino switches, and run experiments with measurement data from 4 datasets. We then evaluate their performance using (i) the original metric used in the solution's paper, and (ii) a novel FL metric which evaluates every solution at FL. This FL metric enables us to delve into an extensive analysis of how the solutions perform on flows of different lengths in diverse use cases. Results show that while some solutions perform similarly across use cases and flow sizes, others show inconsistent behaviours that we discuss.

OBJECTIVE: To determine whether area-level Medicare physician fees for mastectomy and breast conserving surgery were associated with treatment received by Medicare beneficiaries with localized breast cancer and to compare these results with an earlier analysis conducted using small areas (three-digit zip codes) as the unit of observation. DATA SOURCE: Medicare claims and physician survey data for a national sample of elderly (aged 67 or older) Medicare beneficiaries with localized breast cancer treated in 1994 (unweighted n = 1,787). STUDY DESIGN: Multinomial logistic regression analysis was used to estimate a model of treatment received as a function of Medicare fees, controlling for other area economic factors, patient demographic and clinical characteristics, physician experience, and region. PRINCIPAL FINDINGS: In 1994, average Medicare fees (adjusted for the effects of modifiers and procedure mix) for mastectomy (MST) and breast conserving surgery (BCS) were 904 dollars and 305 dollars, respectively. Holding other fees and factors fixed, a 10 percent increase in the BCS fee increased the odds of breast conserving surgery with radiation therapy relative to mastectomy to 1.34 (p = 0.02), while a 10 percent decrease in the MST fee increased the odds of breast conserving surgery with radiation therapy to 1.86 (p < 0.01). CONCLUSIONS: Among older women with localized breast cancer, financial incentives appear to influence the use of mastectomy and breast conserving surgery with radiation therapy. This finding is consistent with the hypothesis that physicians are responsive to financial incentives when the alternative procedures have clinically equivalent outcomes and the patient's clinical condition does not dominate the treatment choice. We also find that the fee effects derived from this analysis of individual data with more precise measurement of both diagnosis and treatment are qualitatively similar to the results of the small-area analysis. This suggests that the earlier study was not severely affected by ecological bias or other data limitations inherent in Medicare claims data.

Oral verrucous carcinoma is a low-grade form of squamous cell carcinoma that commonly appears on the skin and oral mucosa. Clinically, it appears as an ulcer proliferative lesion on the buccal mucosa, further seen on other sites like the gingiva, tongue, and hard palate, or as a proliferative or cauliflower-like lesion [1]. Due to its distinct biologic behaviour, it is a separate entity from oral cavity SCC [2]. VC generally affects people who chew tobacco, smoke, or use snuff. In 1948, Ackerman was the first to define this lesion as a distinct entity [3].

This study explores various strategies employed by YouTube streamer DEANKT to shape audience reception and engagement through live streaming. Using a qualitative approach and reception analysis methodology, this research examines how these differentiated tactics—such as content segmentation, real-time audience interaction via live chats and donations (Saweria), and the implementation of membership tiers—distinctly compare to other channels and influence the dynamic relationship between the streamer and viewers. Observational data were collected from DEANKT's live streaming sessions, complemented by interviews with five active participants and an analysis of community interactions on platforms such as Discord. The findings highlight the effectiveness of DEANKT's diverse strategies in fostering audience loyalty, enhancing viewer participation, and creating a strong sense of community. Key results include the role of membership features in strengthening audience loyalty, the impact of donation-based interactions on viewer satisfaction, and the importance of diverse content segmentation in maintaining audience interest. Most respondents aligned with DEANKT's communication patterns, with the majority positioned in a dominant reception stance, indicating high levels of agreement and engagement with the streamer's strategies. This study contributes to the understanding of how personalized and interactive approaches can transform live streaming into an immersive and participatory experience, ultimately shaping the evolving landscape of digital media consumption.