Dataset A: This dataset contains input/output data generated from a noisy polynomial model: y = f(x)...
Fantastic news! We've Found the answer you've been seeking!
Question:
Transcribed Image Text:
Dataset A: This dataset contains input/output data generated from a noisy polynomial model: y = f(x) + , where x 1 = [2,2] and K f(x) = m(x,a) = px(x; a):= [: m=0 xmam+1. Here, K 1 denotes the polynomial order and a RK+ represents the parameter vector. The dataset consists of two files: SYNDTR.csv, which contains 200 training data points, and SYNDTE.csv, which contains 1800 testing data points. Dataset B: Each file has three columns: The first column represents the input (denoted as TRI for SYNDTR and TEI for SYNDTE). The second column signifies the output under low noise conditions (denoted as TROL for SYNDTR and TEOL for SYNDTE). The third column signifies the output under high noise conditions (denoted as TROH for SYNDTR and TEOH for SYNDTE). This dataset contains real-world data related to fish measurements. Specifically, the input x is 2-dimensional, where x signifies fish width and x signifies fish length. The output y is the fish weight. The dataset is divided into two files: FISHDTR.csv, which contains the training data, and FISHDTE.csv, which contains the testing data. Each file consists of three columns: "width," "length," and "weight." "Width" and "length" serve as the input features (denoted as TRI for FISHDTR and TEI for FISHDTE), while "weight" is the output (denoted as TRO for FISHDTR and TEO for FISHDTE). III. Problems Problem 1 (20%) This problem focuses on Dataset A, initially considering the low-noise (LN) scenario. Assume that you are aware that m is an M-order polynomial, denoted p, for some M. However, the exact order is unknown. Your objective is to train such that f(x) = PM(x, ) approximates the true f(x), utilizing the available training examples for regression. 1. Implement in Python: Create a function "my_model_train" that takes TRI, TROL, and M as input parameters and minimizes train-MSE to return the trained parameters . Your function should employ SVD-based single-shot LS solution. Do not use any prebuilt poly-fit functions. Implement in Python: Create a function "my_model_test" that accepts , TEI, and TEOL as inputs and calculates and returns test-MSE. 2. Implement in Python: Conduct an experiment using the aforementioned functions to compute train-MSE and test-MSE for the optimized with M = 0,1,...,10. In Fig. 1 of your report plot train-MSE (red curve) and test-MSE (blue curve) vs. M = 0,1,..., 10. Discuss how train-MSE and test- MSE vary across M in Fig. 1. 3. Implement in Python: Repeat steps 2-6 for the high-noise (HN) scenario, using the data sets TRI, TEI, TROH, and TEOH. In Fig. 2 of your report, plot train-MSE (red curve) and test-MSE (blue curve) against M. Discuss how train-MSE and test-MSE vary across M in Fig. 2 and compare these observations with those in Fig. 1. 4. Implement in Python: Repeat steps 2-8, this time limiting the training set to the first 180 data points. Create corresponding Figs. 3 and 4 and discuss the results, comparing Figs. 2 and 3 with Figs. 1 and 2. Problem 2 (20%) - Feature/Input/Output Correlations This problem is based on Dataset B. 1. Implement in Python: Utilize TRI and TRO to compute matrix C. For each (i,j) {1,2}, C(i,j) should contain the absolute value of the correlation coefficient between input features i and j. Additionally, C(i, 3) = C(3,i) should contain the absolute value of the correlation coefficient between 3 input feature i and the output. Present matrix C. Discuss the observed correlations and identify which input feature appears to be most strongly correlated with the output. 2. Implement in Python: In Fig. 5 of your report, scatter-plot weight vs width for the training data. Then, in Fig. 6 of your report, scatter- plot weight vs length for the training data. Discuss your observations based on Figs. 5 and 6. FISHDTE.csv - Notepad File Edit Format View Help width, length, weight 3.8352,22.0,169.0 1.2558,12.1,12.2 6.018,30.5,514.0 4.3056, 24.0, 290.0 3.1234,16.8,78.0 2.673,16.3,90.0 3.977,36.0,345.0 2.9415,15.7,70.0 1.408,7.5,5.9 3.555,19.0,110.0 5.1373,25.2,300.0 3.723,22.0,225.0 6.7473,37.4,975.0 5.1338,30.9,610.0 7.1064,36.9,850.0 3.525,20.0,120.0 2.9181,17.5,120.0 3.6636,21.2,200.0 4.02,23.2,242.0 3.825, 22.0,145.0 6.5736,33.7,800.0 3.3957,19.1,110.0 5.5695,31.3,575.0 7.48,59.0,1650.0 7.4624,37.0,1015.0 5.8515,32.7,714.0 4.69,27.6.390.0 FISHDTR.csv- Notepad File Edit Format View Help width, length, weight 2.268,12.9,40.0 4.1272, 22.1,200.0 4.335,25.4,265.0 1.38,10.4,9.7 5.376,44.8,770.0 3.2943,19.4,120.0 2.9044,17.5,78.0 6.8684,39.0,1100.0 2.6316,15.0,51.5 5.2785,26.8,500.0 3.3756,30.0,200.0 3.3516,19.0,0.0 3.624,20.5,150.0 4.896,42.0,500.0 3.6835,23.0,180.0 6.7408,34.0,700.0 2.0672,13.2,19.7 3.525,20.0,130.0 1.16,10.0,7.5 5.7276,31.0,650.0 7.7957,32.5,840.0 5.1042,28.4,475.0 5.3704, 31.4,685.0 6.3705,38.0,950.0 4.7736,25.0,272.0 6.1306,31.8,680.0 3.534.19.3.130.0 SYNDTR.csv - Notepad File Edit Format View Help TRI, TROL, TROH -2.0,0.7797801013578294, 3.110129736916998 -1.979899497487437,0.7334900940715499,3.522017571464012 -1.9597989949748744,0.6873845575640217,-9.401158212267045 -1.9396984924623115,0.6643223335102355, -3.7010739324007367 -1.9195979899497488,0.6045791321175524,-0.9158959772170008 -1.899497487437186,0.5863989332557281,3.837844681297914 -1.879396984924623,0.539376609695645,5.477940776987349 -1.8592964824120604, 0.47669480059185687,-2.588725395476113 -1.8391959798994975,0.46761814899616266,1.9280784320432742 -1.8190954773869348,0.4344378076039004, -1.4545775317712915 -1.7989949748743719,0.38993604832884404,7.429099183859774 -1.778894472361809,0.36941409556698585,1.5680393250941669 -1.7587939698492463,0.3247242149039432,1.7626066596825398 -1.7386934673366834,0.30337611332530134,1.9002791800822811 -1.7185929648241207,0.27695382931911905,9.63758375231981 -1.6984924623115578, 0.2417380470333166,5.159201515070963 -1.678391959798995,0.21266488428847954,2.971023707626178 -1.6582914572864322,0.18984121550654917,1.529028142347971 -1.6381909547738693,0.19995888997041456,7.4287738552470275 -1.6180904522613067,0.14080836239678757,3.674814998965255 -1.5979899497487438,0.1448777350804104, -4.205201611525807 -1.5778894472361809,0.09931189882651313,2.718938409054841 -1.557788944723618, 0.0849791192817739, -3.7120414889977074 -1.5376884422110553,0.07223210822406867,0.8380130468712469 -1.5175879396984926,0.05270961076492344,0.4786649558186202 -1.4974874371859297,0.018114822915554447,-1.4007347663089225 -1.4773869346733668,0.008053188708364923.9.606248271593234 SYNDTE.csv - Notepad File Edit Format View Help TEI, TEOL, TEOH -2.0,0.7673065575965415,0.1235565095361627 -1.9977765425236242,0.7943153591932017,1.7326977723474648 -1.9955530850472485,0.7555639265809324,1.6731494031235508 -1.9933296275708727,0.7614462415560284,1.0054139405737343 -1.991106170094497,0.7606488308433215,4.039788952036721 -1.9888827126181212,0.761312626856185, -0.29215073589232987 -1.9866592551417455,0.7489044855478995, -2.182250446781628 -1.9844357976653697,0.7570355170544998,1.7942388306922106 -1.982212340188994,0.7416562199777219,2.8470072452199324 -1.9799888827126182,0.7582756085459741, 4.074939868993058 -1.9777654252362424,0.7395068145658071,2.424837575239192 -1.9755419677598667,0.7366429368356119,10.821674885305438 -1.973318510283491,0.7241662664281812,4.508093019131171 -1.9710950528071152,0.7231835861240143,2.0058279489547886 -1.9688715953307394,0.7120192568545678,3.51913158306964 -1.9666481378543634,0.719804542660765, -7.482037309227453 -1.9644246803779877,0.7044766896317838, -9.241684444324129 -1.962201222901612,0.6948753549479644,0.3376946557104504 -1.9599777654252362,0.6957621694749753,0.10309533071685884 -1.9577543079488604,0.6861108321395015, 5.149506650813418 -1.9555308504724846,0.6813029142267479,0.028858605978796303 -1.953307392996109,0.6893520984180296,3.1650418604655735 -1.9510839355197331,0.6855969508171061,2.1792827212201322 -1.9488604780433574,0.6774649848763484,2.0721530158573676 -1.9466370205669816,0.662126569361612,3.810070166479305 -1.9444135630906059, 0.6511433187121868,2.1129793296285344 -1.9421901056142301,0.6570036100967817.1.2526295448362181 Dataset A: This dataset contains input/output data generated from a noisy polynomial model: y = f(x) + , where x 1 = [2,2] and K f(x) = m(x,a) = px(x; a):= [: m=0 xmam+1. Here, K 1 denotes the polynomial order and a RK+ represents the parameter vector. The dataset consists of two files: SYNDTR.csv, which contains 200 training data points, and SYNDTE.csv, which contains 1800 testing data points. Dataset B: Each file has three columns: The first column represents the input (denoted as TRI for SYNDTR and TEI for SYNDTE). The second column signifies the output under low noise conditions (denoted as TROL for SYNDTR and TEOL for SYNDTE). The third column signifies the output under high noise conditions (denoted as TROH for SYNDTR and TEOH for SYNDTE). This dataset contains real-world data related to fish measurements. Specifically, the input x is 2-dimensional, where x signifies fish width and x signifies fish length. The output y is the fish weight. The dataset is divided into two files: FISHDTR.csv, which contains the training data, and FISHDTE.csv, which contains the testing data. Each file consists of three columns: "width," "length," and "weight." "Width" and "length" serve as the input features (denoted as TRI for FISHDTR and TEI for FISHDTE), while "weight" is the output (denoted as TRO for FISHDTR and TEO for FISHDTE). III. Problems Problem 1 (20%) This problem focuses on Dataset A, initially considering the low-noise (LN) scenario. Assume that you are aware that m is an M-order polynomial, denoted p, for some M. However, the exact order is unknown. Your objective is to train such that f(x) = PM(x, ) approximates the true f(x), utilizing the available training examples for regression. 1. Implement in Python: Create a function "my_model_train" that takes TRI, TROL, and M as input parameters and minimizes train-MSE to return the trained parameters . Your function should employ SVD-based single-shot LS solution. Do not use any prebuilt poly-fit functions. Implement in Python: Create a function "my_model_test" that accepts , TEI, and TEOL as inputs and calculates and returns test-MSE. 2. Implement in Python: Conduct an experiment using the aforementioned functions to compute train-MSE and test-MSE for the optimized with M = 0,1,...,10. In Fig. 1 of your report plot train-MSE (red curve) and test-MSE (blue curve) vs. M = 0,1,..., 10. Discuss how train-MSE and test- MSE vary across M in Fig. 1. 3. Implement in Python: Repeat steps 2-6 for the high-noise (HN) scenario, using the data sets TRI, TEI, TROH, and TEOH. In Fig. 2 of your report, plot train-MSE (red curve) and test-MSE (blue curve) against M. Discuss how train-MSE and test-MSE vary across M in Fig. 2 and compare these observations with those in Fig. 1. 4. Implement in Python: Repeat steps 2-8, this time limiting the training set to the first 180 data points. Create corresponding Figs. 3 and 4 and discuss the results, comparing Figs. 2 and 3 with Figs. 1 and 2. Problem 2 (20%) - Feature/Input/Output Correlations This problem is based on Dataset B. 1. Implement in Python: Utilize TRI and TRO to compute matrix C. For each (i,j) {1,2}, C(i,j) should contain the absolute value of the correlation coefficient between input features i and j. Additionally, C(i, 3) = C(3,i) should contain the absolute value of the correlation coefficient between 3 input feature i and the output. Present matrix C. Discuss the observed correlations and identify which input feature appears to be most strongly correlated with the output. 2. Implement in Python: In Fig. 5 of your report, scatter-plot weight vs width for the training data. Then, in Fig. 6 of your report, scatter- plot weight vs length for the training data. Discuss your observations based on Figs. 5 and 6. FISHDTE.csv - Notepad File Edit Format View Help width, length, weight 3.8352,22.0,169.0 1.2558,12.1,12.2 6.018,30.5,514.0 4.3056, 24.0, 290.0 3.1234,16.8,78.0 2.673,16.3,90.0 3.977,36.0,345.0 2.9415,15.7,70.0 1.408,7.5,5.9 3.555,19.0,110.0 5.1373,25.2,300.0 3.723,22.0,225.0 6.7473,37.4,975.0 5.1338,30.9,610.0 7.1064,36.9,850.0 3.525,20.0,120.0 2.9181,17.5,120.0 3.6636,21.2,200.0 4.02,23.2,242.0 3.825, 22.0,145.0 6.5736,33.7,800.0 3.3957,19.1,110.0 5.5695,31.3,575.0 7.48,59.0,1650.0 7.4624,37.0,1015.0 5.8515,32.7,714.0 4.69,27.6.390.0 FISHDTR.csv- Notepad File Edit Format View Help width, length, weight 2.268,12.9,40.0 4.1272, 22.1,200.0 4.335,25.4,265.0 1.38,10.4,9.7 5.376,44.8,770.0 3.2943,19.4,120.0 2.9044,17.5,78.0 6.8684,39.0,1100.0 2.6316,15.0,51.5 5.2785,26.8,500.0 3.3756,30.0,200.0 3.3516,19.0,0.0 3.624,20.5,150.0 4.896,42.0,500.0 3.6835,23.0,180.0 6.7408,34.0,700.0 2.0672,13.2,19.7 3.525,20.0,130.0 1.16,10.0,7.5 5.7276,31.0,650.0 7.7957,32.5,840.0 5.1042,28.4,475.0 5.3704, 31.4,685.0 6.3705,38.0,950.0 4.7736,25.0,272.0 6.1306,31.8,680.0 3.534.19.3.130.0 SYNDTR.csv - Notepad File Edit Format View Help TRI, TROL, TROH -2.0,0.7797801013578294, 3.110129736916998 -1.979899497487437,0.7334900940715499,3.522017571464012 -1.9597989949748744,0.6873845575640217,-9.401158212267045 -1.9396984924623115,0.6643223335102355, -3.7010739324007367 -1.9195979899497488,0.6045791321175524,-0.9158959772170008 -1.899497487437186,0.5863989332557281,3.837844681297914 -1.879396984924623,0.539376609695645,5.477940776987349 -1.8592964824120604, 0.47669480059185687,-2.588725395476113 -1.8391959798994975,0.46761814899616266,1.9280784320432742 -1.8190954773869348,0.4344378076039004, -1.4545775317712915 -1.7989949748743719,0.38993604832884404,7.429099183859774 -1.778894472361809,0.36941409556698585,1.5680393250941669 -1.7587939698492463,0.3247242149039432,1.7626066596825398 -1.7386934673366834,0.30337611332530134,1.9002791800822811 -1.7185929648241207,0.27695382931911905,9.63758375231981 -1.6984924623115578, 0.2417380470333166,5.159201515070963 -1.678391959798995,0.21266488428847954,2.971023707626178 -1.6582914572864322,0.18984121550654917,1.529028142347971 -1.6381909547738693,0.19995888997041456,7.4287738552470275 -1.6180904522613067,0.14080836239678757,3.674814998965255 -1.5979899497487438,0.1448777350804104, -4.205201611525807 -1.5778894472361809,0.09931189882651313,2.718938409054841 -1.557788944723618, 0.0849791192817739, -3.7120414889977074 -1.5376884422110553,0.07223210822406867,0.8380130468712469 -1.5175879396984926,0.05270961076492344,0.4786649558186202 -1.4974874371859297,0.018114822915554447,-1.4007347663089225 -1.4773869346733668,0.008053188708364923.9.606248271593234 SYNDTE.csv - Notepad File Edit Format View Help TEI, TEOL, TEOH -2.0,0.7673065575965415,0.1235565095361627 -1.9977765425236242,0.7943153591932017,1.7326977723474648 -1.9955530850472485,0.7555639265809324,1.6731494031235508 -1.9933296275708727,0.7614462415560284,1.0054139405737343 -1.991106170094497,0.7606488308433215,4.039788952036721 -1.9888827126181212,0.761312626856185, -0.29215073589232987 -1.9866592551417455,0.7489044855478995, -2.182250446781628 -1.9844357976653697,0.7570355170544998,1.7942388306922106 -1.982212340188994,0.7416562199777219,2.8470072452199324 -1.9799888827126182,0.7582756085459741, 4.074939868993058 -1.9777654252362424,0.7395068145658071,2.424837575239192 -1.9755419677598667,0.7366429368356119,10.821674885305438 -1.973318510283491,0.7241662664281812,4.508093019131171 -1.9710950528071152,0.7231835861240143,2.0058279489547886 -1.9688715953307394,0.7120192568545678,3.51913158306964 -1.9666481378543634,0.719804542660765, -7.482037309227453 -1.9644246803779877,0.7044766896317838, -9.241684444324129 -1.962201222901612,0.6948753549479644,0.3376946557104504 -1.9599777654252362,0.6957621694749753,0.10309533071685884 -1.9577543079488604,0.6861108321395015, 5.149506650813418 -1.9555308504724846,0.6813029142267479,0.028858605978796303 -1.953307392996109,0.6893520984180296,3.1650418604655735 -1.9510839355197331,0.6855969508171061,2.1792827212201322 -1.9488604780433574,0.6774649848763484,2.0721530158573676 -1.9466370205669816,0.662126569361612,3.810070166479305 -1.9444135630906059, 0.6511433187121868,2.1129793296285344 -1.9421901056142301,0.6570036100967817.1.2526295448362181
Expert Answer:
Answer rating: 100% (QA)
Solutions Step 1 Problem 1 Step 1 Implement the mymodeltrain function in Python This function should take TRI TROL and M as input parameters and retur... View the full answer
Related Book For
Posted Date:
Students also viewed these programming questions
-
a) A 5-Stage pipeline is composed of the following stages Instruction Fetch (IF), Decode (DE), Execute (EX), Memory Access (ME) and Register Write-back (WB). Assume the pipeline does not have a...
-
Can you help me change this in c++ update your payroll program to use a C++ class. Each employee class object should hold the master file information for one employee. You can assume that the company...
-
The following additional information is available for the Dr. Ivan and Irene Incisor family from Chapters 1-5. Ivan's grandfather died and left a portfolio of municipal bonds. In 2012, they pay Ivan...
-
Use the limit definition of the definite integral with right Riemann sums and a regular partion to evaluate the following definite integrals. Use the Fundamental Theorem of Calculus to check your...
-
A ball is thrown at 20.0 m/s inside a boxcar moving along the tracks at 40.0 m/s. What is the speed of the ball relative to the ground if the ball is thrown? (a) Forward (b) Backward (c) Out the side...
-
A silver rod and a SHE are dipped into a saturated aqueous solution of silver oxalate, Ag2C2O4, at 25C. The measured potential difference between the rod and the SHE is 0.589 V, the rod being...
-
You rolled a six-sided die 60 times and got the following tally. 20 ones 20 twos 15 threes 3 fours 2 fives 0 sixes Does this seem like a reasonable result? What inference might you draw from the...
-
Refer to the data for the Blending Department of Sunspots Beverages, Ltd., in Problem 4-13. Assume that the company uses the FIFO method rather than the weighted-average method in its process costing...
-
You are a recent hire at a company. Your manager explains that the company has not used budgets previously. She asks you to explain how budgets can be useful to a manufacturing firm. Create a...
-
A $1000, 9.5% bond with semi-annual coupons redeemable at par on March 1, 2026, was purchased on September 19, 2017, to yield 7% compounded semi-annually. What was the purchase price?
-
What psychological mechanisms underlie the sense of belonging and collective identity within communal groups, and how do these dynamics impact individual well-being and societal resilience?
-
What psychological mechanisms underlie the sense of belonging and collective identity within communal groups, and how do these dynamics impact individual well-being and societal resilience?
-
Add as indicated. Write all answers in lowest terms. 3 - + 2 X X-7 3X 2 = X-7 (Simplify your answer.)
-
United Airlines sells a round-trip ticket for a flight from Boston to London for $750. The customer also receives 5,000 award miles, equivalent to 20 percent of the miles required for a free domestic...
-
Behavioral economics concepts apply to many areas of daily life, including marketing, diet, and exercise. Conduct an analysis of ways government and business use behavioral economics techniques to...
-
How old is a piece of wood that has 0.7813% of the carbon-14 of a living organism? Express your answer numerically in years using four significant figures. Do not include a comma in the answer View...
-
How do individual companies respond to economic forces throughout the globe? One way to explore this is to see how well rates of return for stock of individual companies can be explained by stock...
-
Let : R R' be a ring homomorphism and let N be an ideal of R. a. Show that [N] is an ideal of [R]. b. Give an example to show that [N] need not be an ideal of R'. c. Let N' be an ideal either of...
-
Give a basis for the indicated vector space over the field. Q(2) over Q
-
Let a : Z 5 [x] Z 5 be an evaluation homomorphism as in Theorem 22.4. Use Fermat's theorem to evaluate 3 (x 231 + 3x 117 - 2x 53 + 1). Data from Theorem 22.4 (The Evaluation Homomorphisms for...
-
Discuss briefly the objective and requirements of the ISO 14001 scheme and Europe's eco-management and audit scheme (EMAS), and list the key differences in the requirements of the two schemes.
-
When the sum of the \(x\) values is equal to zero, the calculation of the coefficients of the regression line of \(Y\) on \(x\) is greatly simplified; in fact, their estimates are given by...
-
It is tedious to perform a least squares analysis without using a computer. We illustrate here a computer-based analysis using the MINITAB package. The observations on page 328 are entered in...
Study smarter with the SolutionInn App